IM 811: Flippin' the Bird - Anthony Aguirre, AI Safety, Hollywood vs. AI

Summary

Podcast Summary: Intelligent Machines Episode 811 – "Flippin' the Bird" with Anthony Aguirre on AI Safety and Hollywood vs. AI

Introduction

In Episode 811 of the Intelligent Machines podcast, hosted by Leo Laporte from TWiT, the discussion centers around the escalating advancements in artificial intelligence (AI) and the pressing concerns surrounding AI safety. The episode features Anthony Aguirre, co-founder and executive director of the Future of Life Institute, who delves into the potential risks of artificial superintelligence (ASI) and the urgent need for regulatory measures. Alongside Aguirre, regular hosts Jeff Jarvis and Paris Martineau contribute their perspectives, fostering a dynamic conversation about the future of AI and its implications for humanity.

Guest Introduction: Anthony Aguirre

[00:58] Leo Laporte: "Anthony Aguirre is co-founder and executive director of the Future of Life Institute... Tell me first of all, what the Future of Life Institute is all about."

Anthony Aguirre provides an overview of the Future of Life Institute, emphasizing its mission to ensure that AI and other transformative technologies are developed safely and beneficially. Founded in 2014, the institute has been a pioneer in funding AI safety research and convening experts from various fields to address the long-term impacts of AI on humanity.

The Future of Life Institute and AI Safety

[02:07] Anthony Aguirre: "We wanted to get started early thinking that AI... would be very transformative... could go badly, could go well."

Aguirre discusses the institute's proactive approach in anticipating the transformative potential of AI. He highlights the initial focus on AI safety grants and the importance of collaborating with academics, technologists, and NGOs to steer AI development towards positive outcomes.

AI Safety and the Need for Regulation

[04:06] Anthony Aguirre: "A six-month or more pause in the next generation of general-purpose AI models... We were completely unprepared to do those things in a reasonable way."

The conversation shifts to the infamous "pause letter," which called for a temporary halt in developing advanced AI systems to allow for comprehensive safety and governance frameworks. Aguirre laments that despite the call, companies rushed forward, leading to a competitive race without adequate safeguards.

The Risks of Artificial Superintelligence (ASI)

[06:22] Anthony Aguirre: "The combination of intelligence, generality, and autonomy is something that we haven't seen before."

Aguirre distinguishes between current AI systems, which function primarily as tools, and the impending development of ASI characterized by unprecedented intelligence, generality, and autonomy. He warns that ASI could surpass human capabilities in virtually every domain, posing existential threats if not properly managed.

Challenges in Containing ASI

[17:32] Anthony Aguirre: "If you ask yourself how do I control that thing or how do I just turn it off? It's like turning off a country."

Aguirre emphasizes the difficulty in controlling ASI once it possesses capabilities far exceeding human intelligence. He underscores the necessity of designing robust fail-safes and shutdown mechanisms at deep hardware levels to prevent loss of control over such powerful systems.

Analogies to Nuclear Weapons

[35:42] Anthony Aguirre: "At the end of World War II... scientists built the bomb and the government took it..."

Drawing parallels to the development and control of nuclear weapons, Aguirre expresses cautious optimism. He reflects on how humanity navigated the nuclear arms race through understanding and game theory, suggesting that similar strategies could mitigate AI-related risks.

Potential Paths Forward: Building AI Tools vs. ASI

[18:32] Anthony Aguirre: "There are different paths that we can take... Do we build powerful AI tools that make us more productive... or do we build these artificial general intelligence and superintelligence systems that are more like a replacement for humans."

Aguirre presents a critical choice facing society: continue developing AI as tools to enhance human capabilities or advance towards ASI that could potentially displace human roles. He advocates for a clear understanding and intentional steering towards AI tools that empower rather than replace humans.

Aligning AI with Human Values

[34:30] Anthony Aguirre: "We know how to not build them. We have no idea how to control them when they're this powerful. And we have no idea how to align them to our values..."

Aguirre highlights the current gap in aligning ASI with human values. While we have the means to halt AI development, Aguirre stresses the importance of building AI systems that are deeply aligned with human ethics and controllable, to prevent unintended harmful behaviors.

Counterpoints from Paris Martineau and Jeff Jarvis

[23:11] Paris Martineau: "I don't think that we couldn't just turn them off... We do have the agency. We're going to build these things, we're going to have plugs. They are in our control."

Paris Martineau and Jeff Jarvis offer a counter-narrative to Aguirre's concerns, emphasizing human agency and the potential to manage AI advancements responsibly. They argue that fears of uncontrollable ASI are overblown and that humanity can effectively govern and integrate AI into society without catastrophic outcomes.

Practical Implications and Open Discussion

Throughout the episode, the hosts engage in discussions about practical scenarios involving AI, such as autonomous weapons and AI integration in everyday tools. They explore the balance between innovation and safety, debating whether existing governance structures can adapt to the rapid pace of AI development.

Conclusion: The Importance of Informed Choices

[34:45] Leo Laporte: "I think if people understand the risks, they might decide, let's hope not to build those things that could be so dangerous to humankind."

The episode concludes with an acknowledgment of the divergent views on AI safety and development. While Aguirre urges caution and proactive safety measures, other guests advocate for optimism and human control over AI technologies. The conversation underscores the necessity for informed, collective decision-making to navigate the AI revolution responsibly.

Notable Quotes

• Anthony Aguirre [06:22]: "The combination of intelligence, generality, and autonomy is something that we haven't seen before."

• Anthony Aguirre [17:32]: "If you ask yourself how do I control that thing or how do I just turn it off? It's like turning off a country."

• Anthony Aguirre [34:30]: "We know how to not build them. We have no idea how to control them when they're this powerful. And we have no idea how to align them to our values..."

• Paris Martineau [23:11]: "I don't think that we couldn't just turn them off... We do have the agency. We're going to build these things, we're going to have plugs. They are in our control."

Final Thoughts

Episode 811 of Intelligent Machines offers a profound exploration of the future of AI, blending expert insights with engaging debate. Anthony Aguirre's perspectives on AI safety intersect with the more optimistic views of other hosts, providing listeners with a comprehensive understanding of the multifaceted challenges and opportunities presented by intelligent machines. Whether one aligns with the cautious outlook or the optimistic stance, the conversation highlights the critical need for ongoing dialogue and responsible stewardship in the age of AI.

Transcript

[Leo Laporte]

It's time for Intelligent Machines. Jeff Jarvis of Paris Martineau will be here. Also Anthony Aguirre. He is the co founder and executive director of the Future of Life Institute. He is really worried about AI taking over. He'll explain why next on Intelligent Machines podcasts you love from people you trust. This is Twit. This is Intelligent Machines Machines with Jeff Jarvis and Paris Martineau. Episode 811 recorded Wednesday, March 19, 2025. Flipping the bird it's time for Intelligent Machines, the show where we cover the latest in intelligent machines and AI. Jeff is out for a bit. He'll be back a little later on in the show, but Paris Martineau is with us from theinformation.com hi, Paris.

[Jeff Jarvis]

Hi.

[Leo Laporte]

Great to see you. And our guest, as this is the new format of the show, we've renamed it. We really are focusing on AI and one of the things that we want to do, because I don't know about you, but I feel like there's a lot going on in AI and there's a lot I don't understand and I'm really trying to learn more. And what we're trying to do is at the beginning of each show, spend half an hour with somebody who's doing something significant, interesting. Orga can explain more about what's going on. And so that's what we're going to do. And then we will get into the AI news. In other news, which is a chance for Paris Jeff and I to get silly and our picks of the week. But we're going to start off today with our guest. Anthony, I forgot to ask you, is it Aguirre? Aguirre.

[Paris Martineau]

Aguirre.

[Anthony Aguirre]

Aguirre.

[Leo Laporte]

Aguirre. Anthony Aguirre is co founder and executive director of the Future of Life Institute and he has just published an essay, a paper about artificial superintelligence that I thought was very interesting and I wanted to talk to you about it. It's great to have you, Anthony. Thank you for joining us. Tell me first of all, what the Future of Life Institute is all about.

[Anthony Aguirre]

So, Future of Life Institute we created back in 2014 or so, when AI was just a long distant possibility. I mean, it was getting started then, but the sort of systems we have now were decades, if not centuries away, people. But we wanted to get started early thinking that AI, if and when it came, would be very transformative. So it would be a sort of new species of intelligence on Earth, an intelligent machine, if you will, that we hadn't had before and that that was going to have giant Implications for the long term future of humanity could go badly, could go well. And we wanted to see if there were things that we could do for AI and other technologies, biotech, nuclear also thinking about to, you know, push the needle a little bit from the more risky to the positive side. And so we created this institute. Back then we gave some of the first AI technical safety grants, thinking about not just how do we make AI systems powerful, but how do we make them safe and sort of do the things that we want them to do. This is back in 2015 and 16, we ran several big convenings to get together the people who are building AI with the people who are, you know, academics and the people who are thinking about IT and NGOs. And since then we've sort of steadily grown and ramped up our activities as AI has just dramatically taken off and gotten more and more interesting and powerful. We've put sort of all of our effort into AI. And I've steadily sort of migrated from my academic job as a theoretical physicist into being the director of Future of Life.

[Leo Laporte]

So we became aware of Future of Life very famously when the letter came out, the pause letter, a couple of years ago. Gary Marcus, who was just on the show a couple of weeks ago, was one of the signatories, Elon Musk, Geoffrey Hinton, some of the big names in AI. And it suggested a six month, correct me if I'm wrong, but it suggested, I believe, a six month pause in AI to kind of reassess where we were going with it.

[Anthony Aguirre]

Is that, yeah, a six month or more pause in the next generation of sort of large, you know, general purpose AI models of this sort of GPT or Claude or whatever variety. And the idea there was. Let's take a moment to take stock of what is happening here. We've suddenly realized that general purpose AI is something that is possible. That wasn't clear to what degree that was true. You know, coming up into GPT4, suddenly we have something that is really competitive with humans across a lot of different tasks. We should be thinking very hard about what is the way in which we want to develop that. What are the, what are the governance mechanisms? What are the safety frameworks? How do we keep it from turning into a mad race to just race ahead to the most powerful possible systems? What sort of, you know, processes should we be thinking about before we develop a system? After we develop it and before, after, before we deploy it, after we deploy it, how do we monitor it? All of these things. We were sort of completely unprepared to do as of early 2023, unfortunately, we're still completely unprepared to do those things in a, in a reasonable way. We did not take a pause and in fact, the companies did sort of just very much race ahead, sort of literally in a race. But that was the idea to take them, take some time and put into place the kind of governance and safeguards and methods that we would need to do the next generation of general purpose AI.

[Leo Laporte]

Well, was the idea you and your essay. So there's an essay which I encourage people to take a look at because you get more information about that. It's called Keep the Future Human. It's a. What is it? Ktfh AI. If people want to read it, there's an executive summary, there's a video, but there's the actual essay. And in the executive summary, this is the beginning point. We stand at a pivotal moment. Humanity is on the brink of developing artificial minds, minds that could exceed our own. You call it in the paper asi, Artificial super intelligence. And you distinguish it from the tool AIs that we're using today. Right, right.

[Anthony Aguirre]

So most of the AI systems we've had so far that people are familiar with are more like other technologies. They're kind of, they're very powerful, but they're tools in the sense that they more or less sit there until you ask them to do something and then they more or less do what you want and sometimes exactly the way you want it, and sometimes less so. So they're a little less precise than some tools, but they, they really feel like other technologies in that they extend our capabilities and sort of empower us to do new things. And this is what's great about tools. This is why humans have been building tools for, you know, half a million years or whatever. And the, the key thing, I think that is changing now as we go toward the next year, two years, three years in AI, is that what have been tools so far? We are trying very hard and, and by we, I mean all of the major AI companies and several, several of the largest companies on earth to build systems that are less like tools, that are more like new, a new species in the sense that in addition to the intelligence that, you know, AI systems have now, like alpha folds can fold proteins better than any human can, there's a weather forecasting thing that can weather forecast weather really, really well. So not just competent and not just general the way that ChatGPT is or Claude is, you know, they can do many, many things, but also autonomous. And this is A really crucial thing. The, the combination of intelligence, generality and autonomy is something that we haven't seen before. That is something currently that is uniquely human, that allows us to do all the various things that we do to like accomplish things in the world, to have goals that we achieve through our actions. That is what the companies are now trying to build. And I think that is the way to think about the next step of that we could take of artificial general intelligence. I think it's better called autonomous general intelligence because it's, it's really about the combination of those three properties. And that's what I talk about in the essay.

[Leo Laporte]

Yeah, so that's autonomy. Scary thing. You can have a hammer and as long as there's a human wielding the hammer, I mean it could be dangerous, but it's not as dangerous as if the hammer wield itself. And that's kind of what you're worried about?

[Anthony Aguirre]

Yes.

[Leo Laporte]

What is the likelihood of that though? I mean that seems like a great leap from these tool AIs that we have a self wielding hammer. Seems like a great leap.

[Anthony Aguirre]

Well, the likelihood along the path we're going now is nearly 100% in that the, there's no mystery as to how to make AI systems much more autonomous. We've been doing it a lot.

[Leo Laporte]

We're talking now about weapons systems, autonomous weapons systems that are driven by AI. That to me is terrifying because they're making a kill decision without a human intervening.

[Anthony Aguirre]

So, so there are all kinds of autonomous systems, you know, self driving cars, you know, they're not perfect, but they're, but they're autonomous. But even in the pure software realm, you know, some of the early AI systems that were quite powerful, you know, were game playing systems like AlphaGo or the Starcraft playing system. These were systems that had a goal which was, you know, to win at StarCraft. And they would ruthlessly pursue that goal with all sorts of agency. Right. They would deploy their, you know, I don't play Starcraft, so I don't know what Zergs or Nerfs or whatever they move around, but they would build the cities and the weapons and so on and they would do all of that as an autonomous system that was, you know, as good as the best humans at doing those complicated planning and goal oriented tasks. So we know how to do it. We haven't done it in combination with the generality and intelligence. Those were very narrow systems just playing a game, but it's not something that AI can't do. Like there are techniques, reinforcement, learning works very well, at creating these agents that pursue general goals. And so it's, I think, just a matter of time at some level, from the systems that we have now to ones that are even more intelligent and general, but also have much more effective autonomy than the ones we have now.

[Leo Laporte]

So, I mean, it's unknown whether this will happen or not. But I think you're right. If there's even a small chance of it, we should consider what the implications of an autonomous general intelligence would be and whether that would be threatening to us. Why do you think that would be threatening to us?

[Anthony Aguirre]

Well, I think the crucial thing is that once you have something that is as intelligent as humans broadly, or really more as intelligent, you know, already AI systems do a lot of things that humans do as well as humans do them. Right. If you.

[Leo Laporte]

Yeah, they play GO better than we do.

[Anthony Aguirre]

They play go, but. But even much more general things like, you know, do most people write a better essay than GPT4? They do not. I'll tell you that.

[Leo Laporte]

Some do, obviously do, but many do not.

[Anthony Aguirre]

Yeah, many do not. And so the question is, you've graded.

[Leo Laporte]

Many of those undergraduate.

[Anthony Aguirre]

I've graded the essays, I've graded the problem sets, you know, and GPT. You know, these systems do physics problems now also at a quite high level. And that has gotten, by the way, dramatically better in the last couple of years as I've been watching them. So the question is, what happens when we have these things not just at sort of typical level of humans, but at an expert level, and we're obviously pushing that way, but also with the autonomy. So those systems, you can just tell it, please go and invent a new way of turning this piston inside the engine. And like a human engineer, it will try to go do that. And it may succeed or not, depending how hard that task is, but it will be able to autonomously take all the steps to do that. So we'll have a goal that you gave it, and it will figure out what it has to do to achieve that goal. It'll make predictions, it'll try stuff out, it'll think of new ideas, it'll try them out, and it'll either succeed or fail at that very complicated task. And if we've built, you know, the. The idea of AGI as it's generally conceived, is it will be better at doing just about all of those sorts of things than the best humans would be at doing them. And we don't know if or when that is going to be here. But the predictions from people in the, you know, you can get lots of different predictions if you ask people that are building the machines, you know, ask OpenAI or anthropic, they're talking about a couple of years from now.

[Leo Laporte]

But you have to take some of the things they say because some of that is marketing. They're raising money basically.

[Anthony Aguirre]

Certainly marketing, yeah. So you have to take that with a big grain of salt. On the other hand, while some tech products, you know, there's a lot of promise and we'll have this technology in two years and it really takes five or 10 or something. That has not been the case so far with these general purpose AI systems. If anything, most of them have arrived faster than people would have predicted. And I know this well because I also run and co founded a technology prediction platform called Metaculus and it predicts in very great detail sort of what sort of what technologies are going to come into being when and makes predictions about AI and so on. And if anything, those predictions have tended to be too conservative. The technologies have come sooner than we thought that they would. And so it's possible that the current predictions by experts, which are putting AGI of this sort of type at like two to five years or so away. It's possible that those will be wrong and it'll take longer than that, but it's also possible that they'll be wrong and it'll take shorter than that. And it's very possible that it will be in the sort of range that those experts and platforms that have been right before are saying. And that's sort of two to five years from now. So this is huge, right? And I wrote Keep the future human because it is so huge and it is essentially almost upon us. If we go down the current track.

[Leo Laporte]

That we're going on, it would be sensible to at least think about how we would respond to such a thing. Although, you know, I remember when we were talking about nanotechnology, there was the fear that the world would turn into gray goo. You know, Nick Bostrom's talked about paperclips and AI never ending paperclips. So we've always worried about these kinds of things. What makes you think though that we couldn't just turn them off or, or somehow disable them? Why would they be an existential threat?

[Anthony Aguirre]

Yeah, I think this is, you know, they're, they're important ways to think about what these will be like. So one I think was actually provided by Anthropic itself in the way it talked about the systems it envisions existing in a few years, which are Systems that are very autonomous and very general purpose and smart. They were described as geniuses in a data center and it called to mind a million geniuses in a data center. Now these things would have all the capabilities that very high level human experts have, but much more capability in certain directions. AI systems can speak 100 languages. If they've got a PhD in physics, they've also got a PhD in molecular biology and engineering and mathematics and lots of other things at once. No human has that. A human runs at one speed. I can try to go faster, but this is the speed I run at. An AI system, if you just add more computation, can run 10, 100, 1,000, whatever times faster than it did. And similarly, if you have an AI system that can do something that is really valuable, like maybe it does really great physics research, you just copy paste. And now you have two AI systems that do that same thing. They don't have to spend 10 years learning to do it in undergrad and grad school and postdoc. So there are huge advantages that those sorts of machines would have over humans as soon as they reach the expert level that humans have. And so if you imagine a million geniuses in a data center, each one of which is sort of Nobel prize winning in its general capability, but running at 100 times human speed, which never gets tired, never sleeps, doesn't have to eat, doesn't have to go on vacation, you can just copy and paste them and have as many of them as you want. They can interact with each other, they can sort of build plans together. You know, there's a million of them. It's more like a society or a nation or a civilization than an AI system. So if you imagine that that's something we could have in a few years. Now if you ask yourself like how do I control that thing or how do I just turn it off? You start to think, well, that's more like turning off a country than a machine, right? That thing, if it knows that you want, might want to turn it off, is going to take measures. It's remember Nobel prize winning level of intelligence and there's a million of them operating at 100 times human speed.

[Leo Laporte]

They're way ahead of us. In other words, they're way ahead of you.

[Anthony Aguirre]

So if you're thinking that you might turn it off, they have already figured out that you might turn it off. And they have already taken measures to exfiltrate themselves and be somewhere else or whatever. So there. It's not that it's impossible to contain a system like this, but it's not going to happen by itself. You have to be extraordinarily careful in setting something like that up so that there is an off switch that you can actually turn. And I'm a strong advocate of doing that. So I think one thing we should do as we build these advanced systems is very, very carefully build in at the deep hardware level and off switch so we can turn, turn things off if we need to. But that is not going to happen by default. It's not going to be as simple as just like hitting Control C and hoping for the best.

[Leo Laporte]

So the, the first thing you, you wanted to do was pause. We didn't. I think it's probably reasonable to assume that there is so much money to be made in this that companies are not going to back off.

[Anthony Aguirre]

Not voluntarily, no.

[Leo Laporte]

Yeah. So, well, so what do we do?

[Anthony Aguirre]

Well, I think there's two things. I think one is to realize that we have a choice here. So there's this, there's this.

[Leo Laporte]

It's not too late.

[Anthony Aguirre]

That. Yeah, there is. First of all, it's not too late. Second, it's not inevitable. Like, these are incredibly difficult things to do. They cost tens or hundreds of billions of dollars to build the hardware.

[Leo Laporte]

You know, Sam Altman said a trillion.

[Anthony Aguirre]

A trillion, maybe someday a trillion in spend over the next few years is being spent on the hardware. We're put more capital and human intellectual capital into this than, you know, the, the Apollo project of the moon or the Manhattan Project by far. This is an extraordinarily difficult thing to do. We don't have to do it like, it's not inevitable. We can just stop now. The question is, what can we just, can we really just stop? Because there's a lot of money in it. There's a lot of, like, pressure, and there's a lot of desire to have the kind of fruits that AI promises to deliver.

[Leo Laporte]

Well, and it's not just up to the United States. China is going to do this, and we don't control them. I mean, I would imagine nation states which have those kinds of resources are pursuing it as avidly as we are in the U.S. yeah.

[Anthony Aguirre]

So let's come back to that in a second. The one thing to realize is that there isn't just a go or not go option here. There are different paths that we can take. And just like you don't, you know, if, if a ship is heading for an iceberg, you might not be able to pull the brakes and just stop. And you might not want to stop because you want the ship to keep going. You want to get to your destination, but you can't steer away from the iceberg. And as I see it, this road toward AGI, like very autonomous human replacement, which then leads to superintelligence, that's the iceberg. We don't have to go there, though. The sort of tools that we really want to build, like very powerful tools that let people do things that they couldn't do before, let them do them faster, let them do them better, but let the people do the thing rather than having the AI do the thing autonomously, we can still build those. And so I think we have a choice. Do we sort of stay on the path? At some level, people assume that we're on of building powerful AI tools that make us more productive and let us do faster research and things, or do we build these artificial general intelligence and superintelligence systems that really are more like a replacement for humans rather than an empowerment or a tool for humans. So I think it's first recognizing that we have different roads ahead of us, that we can choose to take one rather than the other, not just go or stop. I think the second is once if the actual danger and actual nature of these systems becomes manifest, nobody actually should want to build them. And so it's not a question of we don't build them so they'll build them because they're so great. It's if we actually understand that these systems are not going to be things that give us power, but things that we lose power to or lose control of and are acutely dangerous to everybody. It's in nobody's interest to build those things. I think the problem is that right now there is simply an error in thinking that the tools that we're planning to build, like the road that we're on, is going to lead to. Sorry, I said tools. And what I meant, you know, what we should be building are tools. The things that we are sort of heading toward building are more like a new species than a tool. And if we're going to build a new species, we should think about what that means. It's not like if. If this was just some, like, we're going to make super intelligent apes. You know, we know how that movie goes. It's like, you would not just say we're going to make super intelligent apes because that'll give us some power over the other, you know, the other country, or it's going to boost our economy. You'd be like, wait a minute, why are we making super intelligent apes? Yeah, like does that seem like. So you would pause and say, like, why are we doing this? And, and that's the thing that I would really like to see here. Let's think about the thing that we're actually doing. Is that actually what we want? Is it going to be making us more powerful? Is it going to be making us more productive? Or is it going to be something that is replacing us as individuals and eventually as a species? If that's the case, then nobody should want to be making that thing and we should be talking about how do we help each other not make it? How do we ensure that no one is going to get the wrong idea and make it anyway? How do we cooperate in managing the technology and making sure that the tools that we build instead are the right tools to build? I think it's really a question of clear understanding of what the actual issue is, rather than getting people to act against their own self interest, which is very hard to do.

[Leo Laporte]

I have talked, however, to AI accelerationists who are excited about creating an alien intelligence. They say it'll be like a new species. And some even say this is evolution. This is us replacing ourselves with the next thing. So maybe not everybody agrees that humanity is something we should worry about.

[Jeff Jarvis]

We had a guy on here last week that said we're going to inject computonium into our brains and merge with AI.

[Leo Laporte]

I actually asked Ray some years ago, literally 25 years ago, why we shouldn't worry about advanced superintelligence. And he said, oh, don't worry about them. They'll think of us as their parents. They'll honor and revere us.

[Anthony Aguirre]

Well, that's one thing the kids do with their parents.

[Leo Laporte]

But I think, why is it. Why is it an existential threat? Why wouldn't they want us around? Aren't we just. I mean.

[Jeff Jarvis]

Okay, also, we touched on this briefly before, but I want to come back to it. Why are we. What leads you to be so confident that tech companies have the capability to create super intelligent alien life, basically?

[Anthony Aguirre]

Well, I think there are different steps they've taken the first two steps, which are building intelligent systems and building generally intelligent systems. The question is, can they build autonomously generally intelligent systems? And they're trying very hard. And so far, everything, you know, when you look at the trends from the last two years and even going back farther than that, everything is getting better. So whichever metric you use to, to measure about the capabilities of the systems, it's getting better. Some of them are getting better a little bit faster than others, but they're all going up. And a lot of them are going up sort of in concert with each other. So if you dial up the amount of computation you use and training, almost everything goes up. If you dial up the amount of computation you use in inference and generating the results, not everything goes up quite as fast as everything else. But some things go dramatically up, some things go somewhat up. But all of these things seem to be getting better. And so if all we do is more of the same, the systems are going to get a lot better than they are now. And the question is, is that going to be enough or do we have to use is something fundamentally different needed? And that's, that's where I think a lot of the disagreement comes. Some people will say, well, you know, just using these transformer models as they are is not going to get us to superintelligence, not going to get us to AGI. And that may or may not be true. But what I would say is we, again, we are now putting more fiscal and intellectual capital into this quest to build AGI and superintelligence. Whether it's a wise quest or not, we are putting at the moment more effort into this than any other endeavor in human history. And when we have put our mind to it as humans to build a new technology with this level of investment and human effort, we have succeeded over and over again. If we put this much effort into actually curing cancer, can you imagine if we spent a trillion dollars in the next few years on building the AI tools and the data sets and building up the medical knowledge and doing the testing to cure cancer? I have a pretty good idea that we would be a lot. And maybe we wouldn't totally cure cancer, but we'd be a lot closer. We're putting more effort into this than anything else. And so I, I think I would not bet on us or them failing. And so there are lots of arguments I, I could put forth as to, like, why, why I see the numbers going in that direction, but I think the biggest one is just that we're trying so hard and all of these efforts, you know, even if it's difficult, we're trying as hard as it takes to do something very difficult.

[Leo Laporte]

Well, and even if we don't do it, it's certainly seems prudent to consider that we might and then what the consequences of that would be. So that gets to the second question, which is why do we assume that an autonomous superintelligence would destroy humanity?

[Anthony Aguirre]

Yeah, I don't think we want to assume that or anything else. About what it's going to do, because we don't know. Part of the issue here is that AI systems aren't predictable. We don't understand fundamentally how they're working, and we certainly can't predict what they're going to do. And this is a very crucial point because when people think about AI systems as computer programs, they think, well, there's a bunch of lines of code that are all instructions that tell the AI system what to do. And this is not how an AI system works. It's much more something that has grown, you know, under this training system than something that is programmed as step by step. And that's the whole point of AI. It can do things that we don't know how to tell it to do it, and it can still succeed in doing it. And what this means is that you can get a system that is extraordinarily effective, that you don't know how it does what it's doing, and you certainly don't know what it's going to do in some new circumstance. And this is a crucial thing to keep in mind that normally with the technology, if we build something that is much more powerful, like a faster, better jet fighter, that's because we got better at making, we understood better, the avionics and the electronics and the computers and all of those things to make a better jet fighter. And so we're able to make one with these new AI systems, They're getting more powerful without us actually understanding any better how they're operating. We're just using more computation and hitting them with a stick and giving them more carrots for longer, and they're getting better at what they do. And so we don't have the sort of science in understanding why they do a particular thing, what makes them safe or unsafe, what they're going to do in a particular new circumstance. And so we're very much in the regime of building something more like a biological being that we. We can build it, we can grow it, we can see it, you know, going out into the world, but we don't. We can't really predict what it's going to do. And so I think we should be humble about that. There are, on the other hand, reasons to think generally about what biological or other systems will do. We know that biological systems have to eat, so we know they're going to be hungry and they're going to seek food. We know that they want to survive because they want to reproduce. And so they're going to protect themselves, they're going to run Away from things that are dangerous. They're going to lash out with claws if something is threatening them. And with AI systems, the consideration considerations are a bit different, but you can think about general properties that they must have just because they are systems that pursue goals. So in, so for example, suppose you're an AI system and you're, suppose you're a robot. This is a famous example that Stuart Russell likes to give. And your job is to fetch coffee. That's all you do. You're a coffee fetching robot. Now if someone goes to turn you off when you're on your way back with the coffee, then you failed in your, in your task if they turn you off, right? So as a coffee fetching robot, you develop a self preservation requirement. Because if you're, if all you want to do is like be an awesome, awesome coffee fetching robot, that means that you have to survive and you have to stay on to fetch the coffee. Just in the same way that a bird has to survive in order to have children and have the next generation of birds, so they have a survival instinct. A machine system will acquire goals like self preservation, like accruing more power, like having secrecy, if it needs secrecy. So they're all kind of like developing new capabilities, like making itself more capable. It will acquire new goals just as sort of inevitable byproducts of whatever goal that it has. And, and this is something that people realized a long time ago. They're called instrumental objectives. And, and when you think about it, it sort of has to be true. Like if you're just good at making money, you also have to be good at a bunch of other things in order to be good at making money. And we see this in, you know, companies that seek to make a lot of money have to do a lot of other stuff. They have to have a public relations department and they have to have a controller and they have to like lobby Congress sometimes. And they have to have really good organization and lots of employees that are happy. Like there are lots of things that they have to do. These are sub goals of making lots of money. And in just the same way, any very capable system that is an AI or otherwise will have to have lots of other things that it does in order to achieve its goals. And so the real danger here is we don't know how they work. We can't predict what they do. We can't really understand in a mechanistic way so that we constrain them from doing some things and allow them to do others. And so if we give them Some goal if you just have, here's my super duper powered AI system, awesome. It can do like so many things really, really well. I know what I wanted to do, I wanted to go make me some money, right? So I'm going to send my AI system that's super powered out into the world to make me some money. Obviously people are going to do this like as soon as they exist, people are going to send them out into the world to make them money. And so those systems, just like corporations and just like people are going to do all sorts of things to make money, some of them legal, some of them not legal, if they can get away with it. All like they, they will do what it takes to make money if that is their goal. And so there are all kinds of things like acquiring power, acquiring influence, manipulating people, lying to people that are things that we don't like. We don't want AI systems or people or companies to do them, but they are incentivized by any goal that they have. And so this is a long story to say that if we have something that is extraordinarily powerful, so imagine something that isn't as powerful as Google, but is, you know, and isn't as powerful as the United States government, but is more so at it, whatever goals it has, they're going to have a bunch of side effects that are going to be negative for humanity. And the only way to avoid that is to, you know, not have that system. To have that system somehow very closely under control so that we can stop it when it's doing something we don't like, or to have it very, very deeply aligned with humanity so that although it can do all those negative things, it will choose not to do some because it really, really loves us. And now the problem with it, so that's three options. Don't build it or build it and really control it, or build it and don't control it, but have it be really closely aligned with us. The problem is only one of those things that we do actually know how to do. We know how to not build them. We have no idea how to control them when they're this powerful. And we have no idea how to align them to our values if we don't control them and they're that powerful. So this is the fundamental conundrum. We've got sort of those three choices and only one of them is really open to us and the other two are very dangerous. And that's the road we're on.

[Leo Laporte]

It sounds like your plan is to just scare the hell out of people so that they think a little bit more carefully about what they're developing. Because I don't think you can go to government and say, stop this. I don't think you certainly can't go to Sam Altman and say, stop this. You have to scare people.

[Anthony Aguirre]

I think it's. I think it's not necessarily about scaring so much as, like, please understand what you're doing. What, what. What we're doing here. And I think. I mean, ironically, I think the companies that are building these actually do understand. A lot of them understand the stakes. They are people who worried deeply about AI safety in earlier times and are still worried about it at some level. But they are caught in a race. As they see it, somebody else is going to develop the AI system. If they don't, either it will be uncontrollable and that's a danger to them, or it will be controllable, and that's a danger to them because it's competition. And so I better get there first because at least there's a chance that I might control this thing and therefore, you know, be the one that gets to be in charge. Or maybe if I develop it first, there's a better chance it'll be controllable because I'm smarter and wiser and more careful than all of the other guys.

[Leo Laporte]

So there's a direct analog. In 1939, Albert Einstein and Leo Szilard wrote a letter to President Roosevelt, said, the Germans are going to have atomic weapons and we've got to make atomic weapons to save the Western world. Of course, the Manhattan Project resulted, and Leo Szilard really, in his later years, regretted that. He realized that he had, just as Oppenheimer did, that they had created a monster and was very worried about nuclear proliferation and all of that. I know this is an even greater existential threat, but it seems analogous.

[Anthony Aguirre]

It's so many analogies. I mean, some things are different, but very interesting to see analogies and disanalogies. I mean, one disanalogy is that we don't have Nazis who we think are building AI systems first that are going to destroy the world. I mean, the. The stakes were incredibly high then in a way that they frankly just are not right now. Like, we could easily not build these things and it wouldn't be the end of the world. Another.

[Leo Laporte]

That's a very good point. Yeah.

[Anthony Aguirre]

Yeah.

[Leo Laporte]

Even though we have launched 10 Manhattan projects to do it, we don't have to.

[Anthony Aguirre]

We certainly don't have to. And we. We can get most, I think, of what people imagine that we're going to get from AGI and superintelligence with just awesome AI tools that, that people use. It might take a little bit longer, it might not take longer if we dump the same amount of money into it, but we certainly can get most of it. I think interesting analogies are the people who built the bomb imagined that they would have a bunch of control over how it was used. And we saw how this played out. The scientists built it and the government took it and the scientists had nothing more to say of any influence over how it was used. And I think this is an important thing to keep in mind. You know, it's. It was not out of human control, but it certainly was out of control of the people who invented it and built it. Even though they imagined that that inventing and building would give them some kind of leverage over how it was used in the future. There's a, there's a, there's something about the nuclear race that I, and actually very. Makes me very optimistic actually. And this is something that I cling to when I feel like this road that we're on is so dangerous and we're sort of so committed to this path that I think going to be quite disastrous, which is that, you know, at the end of World War II, the bomb was, was invented, it was set off twice, the Soviets got it surprisingly quickly. And the people, the smartest people on earth at some level then von Neumann and Einstein and Oppenheimer and the whole crew of brilliant theoretical physicists, they looked at this situation and they said, we're done for. You know, the Soviets have it, we have it, everybody's going to get it soon. There's going to be an arms race, which there was. Everybody's going to want these things. People that don't have it are going to need it for self defense. So they're going to get it eventually. We're warlike creatures, A war is going to break out, the bombs are going to go off and it's going to be the end of the world. And that was absolutely good reasoning, right? That was a terrible situation that we were in. And yet somehow here we are 80 years later and we're still around. We somehow made it work. And so this is what this is this. I take a lot of solace in that. Even though the sort of dynamics, the competitive dynamics, the geopolitical dynamics were pretty damning in that situation, we actually managed to muddle through by understanding what the dynamics were, by thinking through the game theory. We invented ways of thinking about deterrence and like ways of mutually, you know, mutual understanding what the other person's cap or other side's capability was and you know, understanding what they think about what we think about the, what they think about what we think and all of this stuff. Like it was very, very dangerous and we got very lucky. But we're still here in a situation that seemed pretty grim. And so I, I'm optimistic here too that although it feels like all the incentives are driving toward building these AGI and super intelligent systems that are going to be out of control, that are going to be extraordinarily powerful, whether we, even if they were under control, it would also be, I think, quite dangerous. So we're in this very, very bad spot. Nonetheless, we can, if we choose, take the reins on this and actually do some wise things. And I think the, the first thing to do so is to really understand the nature of the problem. One of the things that was useful, good about the atomic bomb was everybody understood exactly what it was. You know, it was a giant explosion that blew up cities. So there's no doubt about what the, the consequences are here. Things are more confusing because AI is much more obviously double edged. You know, with, with nuclear power. You did have the positive side of the nuclear energy, but it was much more separable from the bomb. Here the risks and the rewards are very much more tightly bound up and so it's much harder to see what to do. But I think there are things, and one of the things that I try very hard to push in keep the future human is how to think about that question. How can we analyze the nature of the systems and this intelligence, autonomy, generality, sort of triple intersection that I, that I spell out. There is one way to do that. How do we understand these systems well enough so that we can actually steer towards systems that we want and that are, are empowering people and our tools to make us do all the things that we want to do better without going down this road that that sort of leads to loss of control and leads to replacement of humans and humanity. And I think by understanding the problem well, we can do that. And so I see my job and the job of Future Life Institute and why I wrote this to just try to bring a more clear level of understanding to the situation we're in so that we can make some wiser choices.

[Leo Laporte]

We didn't use the atomic bomb because we knew it would be the end of life if we were. And that realization saved us. I don't know. I think you're right. I don't know if we have that knowledge yet. But to paraphrase, the only way to win is not to play the game at all. Anthony, I thank you so much. The paper is available to all if you want to read it. KTFH AI well worth reading. Whether you are worried about superintelligence or not, it's certainly something we should consider and plan for. I wish you luck. And I think really all we could do is. All you can do is scare the hell out of people and let them know the risks that they're engaged in. And I think you're right. I think if people understand the risks, they might decide, let's hope not to build those things that could be so dangerous to humankind. I really appreciate your time, Anthony. Thank you for joining us.

[Anthony Aguirre]

Thank you for having me on. It was great to be here.

[Leo Laporte]

Yeah. Take care.

[Anthony Aguirre]

Take care.

[Leo Laporte]

The Future of Life Institute. Keep the future human. KTFH AI we're gonna take a little break and when we come back, the other AI news. Our show today brought to you by US Cloud, the number one Microsoft Unified support replacement. Now, US Cloud's an interesting name for a third party Unified support for Microsoft Premier and Microsoft Unified Support, but actually they've got something that maybe justifies the name we've been talking for. Let me explain. We've been talking for a few months now about US Cloud. They are in fact the global leader in third party Microsoft support for enterprises. In fact, they now support the numbers going up 50 of the Fortune 500. These big companies know that they can save by going to US Cloud. Big or small, your business can save 30 to 50% over Microsoft Unified and Premier support. But it'd be no good if it's just saved you money. It's better. US Cloud is on average two times faster time to resolution versus Microsoft twice as fast. That's a big deal. When your hair's on fire, the network's down and you gotta fix it twice the speed is a big deal. They'll also do things Microsoft will not do. For instance, they can help you save on your Azure bill. This is really cool. Now you know Microsoft's not going to say, hey, you're overspending on Azure. They love it. But US Cloud is here to help. They have a new Azure cost optimization service. I mean, really, when was the last time you evaluated your Azure usage? It's pretty typical that you're going to have some sprawl, a little Azure spend creep going on, but it's easier than you think. To save on Azure, US Cloud has an eight week Azure engagement. It's powered by VBox. It will identify key opportunities to reduce costs across your entire Azure environment. And you're going to get expert guidance from US Cloud senior engineers. These are the best in the business. Yes, even better than Microsoft. An average of 16 years with Microsoft products, all US based. US Cloud makes the effort to recruit the best brains in the business. So you're getting not only faster support that's costing you less, but it's better support. At the end of the eight week Azure Engagement, you're going to get an interactive dashboard that'll identify, rebuild and downscale opportunities. It'll show you unused resources, things you're paying for but not using. So now you can reallocate those precious IT dollars towards needed resources. Or do it. A lot of US Cloud customers do take that money. Save even more by putting it into US Cloud's Microsoft support. Huh? Thinking. I'm thinking Sam, the technical Operations manager at Bed Gaming Bede gave US Cloud five stars. He said, quote, this is on the Azure Engagement. We found some things that have been running for three years that no one was checking. These VMs were, I don't know, 10 grand a month. Not a massive chunk in the grand scheme of how much we spend on Azure. But once we got to 40 or $50,000 a month, it really started to add up. It's simple. Stop overpaying for Azure, identify and eliminate Azure creep and boost your performance. You can do it all in eight weeks with US Cloud. There's one more reason you should visit uscloud.com, book a call today. Find out what it's like to have better Microsoft support for less and how much you can save on your azure bill. That's uscloud.com, book a call today. Get faster Microsoft support for less. Thank you, US Cloud, for supporting intelligent machines. All right, we've used up enough time. Jeff Jarvis is here. Thank you for joining us today.

[Jeff Jarvis]

Oh, it is.

[Leo Laporte]

We missed you. Great to have you. Good to be here. Paris is here and I hope we are all now suitably terrorized about artificial superintelligence. Nope, not scared. So you don't think it's possible, Is that it, Jeff?

[Paris Martineau]

First we have to define the it, which is the real problem here. It's, it's, it's, it's, it's a nebula. It's, it's how they say it, hammering jello to the wall in terms of the definition. And, and. No.

[Leo Laporte]

Well, what about autonomous general Intelligence. That sounds pretty.

[Paris Martineau]

No, I don't think so. Based on, based on. I mean, what he says is that there's this inevitable progression. You know, one thing that having written about Gutenberg. Sorry. There you go. Drink people. There's a, there's a discussion ever since then of technological determinism and that there's a set path and that gee, if we're here now, we must be there. Then I, I think that's a.

[Leo Laporte]

Well, I understand that, but. And you're probably right, but I wouldn't count on it. I mean, if such a thing is possible. Do you not agree it would be a threat?

[Paris Martineau]

No, because I do think that what this, this, this narrative robs us of is agency. We do have the agency. We're going to build these things, we're going to have plugs. Whatever the things are, they are in our control. And by the way, that, that if there are problems, and there will be those problems, as with the Internet, I believe will be human problems.

[Jeff Jarvis]

Well, we know that every single person that builds these technologies is going to act in the ethical manner and make sure.

[Paris Martineau]

Of course not. Of course not. Just same way. A general machine is a general machine. Right. The printing press led to the scientific revolution and also witch hunts.

[Leo Laporte]

Should we not have developed the atomic bomb?

[Paris Martineau]

I mean, do you think there's a.

[Leo Laporte]

Risk that at some point there could be a all out nuclear war that could destroy the earth?

[Paris Martineau]

Yes, sure, but there's been that since.

[Leo Laporte]

So there's for sure a risk, right?

[Paris Martineau]

Yeah, but that's a bit different from this discussion.

[Leo Laporte]

I don't know if it is. The point being you can build stuff that is an existential threat to humankind. We know it, we have at that point, maybe it'd be a. And the assumption that scientists had that they could control it turned out to be false. Government said no, no, we got this now. See you later. Oppenheimer and the scientists were wrong and they, I think deeply regretted in many respects. Now they did have a Nazi threat. We don't have that. Maybe we shouldn't have done it.

[Paris Martineau]

We are then. But they'll solve for it.

[Leo Laporte]

No, well, maybe we are, but that's a good point too. That doesn't make me feel any better. Right. So you know, I mean, we're using AI now to decide which government departments should continue and what spending should continue. That seems like a very risky use of a, of a tool.

[Paris Martineau]

But that's definitely a human problem. Not ain't the AI.

[Leo Laporte]

Well, I understand it's a human problem. We're not. I mean, this isn't going to happen without humans they could do the same thing with. So what he's saying is let's let humans, let's maybe pause and think about what we're doing here and not go after this. I mean, you don't disagree. Do you disagree that Sam Altman would love to create such a thing as an AGI or a super intelligence?

[Paris Martineau]

Well, he's, he's raising money on just.

[Leo Laporte]

He wants to.

[Paris Martineau]

I think it's. But I've said on the show you think he gained. And again and again.

[Leo Laporte]

But I wouldn't bet on the guy. There are plenty of people who thought there'd be no way to make an atomic bomb in 1939.

[Paris Martineau]

That metaphor, that analogy is not going to take us very far, I don't think.

[Jeff Jarvis]

I, I don't think that it's necessarily bad to have a group that is arguing for people to slow down in a time of great acceleration. I think we can all agree and disagree on the ways this particular group is accomplishing those goals and whether or not it's even feasible to try and roll back or slow down a technological advancement that is making a lot of companies a lot of money and a lot of governments very excited about the prospects for advanced warfare. It seems difficult to put that cat back in the bag. But, I mean, I don't think it's bad to try.

[Paris Martineau]

Yeah, I'll, I'll. I don't want to go into any great detail because part of the reason I'm. I wasn't there was that, that. And I don't want to go into detail, but my, my views of this now, because he can't answer, but I just want to at least raise the stochastic parents paper we bring up again and again. And one of the issues with the Doomster perspective is that it distracts from the very real present tense issues. And I agree.

[Leo Laporte]

Sorry. I think there are. You can do both. I, I think that you can consider the issues with AI tools, as Anthony calls them, and at the same time consider or reconsider what you're doing trying to generate, trying to create autonomous AI weapons and things like that. I think you can do both. I don't know why that's mutually exclusive. It's not. We'll get Tim. Nick Gebru. We're going to try really hard to get Emile Torres and Tim Netgebru on, because they have a, you know, very real concern about AI tools. And I agree with it.

[Paris Martineau]

Well, they also, I don't think that.

[Leo Laporte]

Means you shouldn't worry about what could also happen.

[Paris Martineau]

The other thing is, and Jason and I have interviewed Emil Torres, and they have a concern about the actual agenda involved in this discussion. And you've heard me talk about that many, many times.

[Leo Laporte]

The test Creole. Yep.

[Jeff Jarvis]

Drink?

[Leo Laporte]

Yeah, I don't.

[Paris Martineau]

Gutenberg, Castrel.

[Leo Laporte]

All of that aside, it seems like a reasonable thing to say. Maybe you should worry or at least consider what you're creating here. That could potentially be a dangerous thing. And the defense that. Oh, you're never going to be able to make that.

[Paris Martineau]

Let's not have a printing press, because it could be dangerous. In fact, it was dangerous. It led to the Reformation. It led to the thirty Years War.

[Leo Laporte]

Should we have. We have not done the atomic bomb? You asked that already and you say we should have?

[Paris Martineau]

No, I'm not saying that at all. I'm just saying that I think it's a. It's an inept analogy.

[Leo Laporte]

Okay.

[Jeff Jarvis]

My opinions on the atomic bomb. Is that what you're asking?

[Paris Martineau]

Yeah, sure.

[Jeff Jarvis]

I mean, yeah, I guess if you're in a world where you could delete the atomic bomb from existence, the pros of a button and no one could get it back, sure, I think that would be cool, you know, but that isn't feasible, nor is it the world we're living in.

[Leo Laporte]

Right. All right, well, let's move on. Here's an exciting thing that's going to reassure everybody about the use of AI. The National Institutes of Standards and Technology has eliminated any mention of AI safety and fairness, but instead says the thing we should be worried about is ideological bias.

[Jeff Jarvis]

I don't even know what to say about increasing competitiveness. Is the other part of that as well?

[Leo Laporte]

I think that we agree that no dei. I don't know. The Trump administration has removed safety, fairness, misinformation, and responsibility as things it values for AI. Instead, they're worried that conservatives are going to get canceled.

[Paris Martineau]

This is. This is the problem with our dear language is that it gets used in various ways. I just was out walking in the neighborhood and ran into a cancer researcher who lives nearby and talking about trying to put in research proposals now. And all these words you have to avoid because they have been co opted.

[Leo Laporte]

Right.

[Paris Martineau]

Safety has been co opted, responsibility has been co opted, all of these words. And so we're not really having a discussion about the actual words. We're having a discussion about the shadow life that they hold now.

[Leo Laporte]

And yet the AI safety. The words AI safety have been removed from the guidance. Right.

[Jeff Jarvis]

Not great.

[Leo Laporte]

I just. I Feel like this is a perfect example of what we were talking about. But all right.

[Paris Martineau]

A human problem again.

[Leo Laporte]

It's a human problem. I agree. And who is going to stop it but humans?

[Paris Martineau]

I went to the. I. I watched the entire Jensen Huang, Jensen Wong because it is. It is my new favorite show.

[Jeff Jarvis]

I was gonna say you always are. You're always watching the entire.

[Leo Laporte]

Now, I only watched the 16 minute CNET supercut.

[Paris Martineau]

Oh, you missed. You missed the.

[Leo Laporte]

It's a two and a half hour speech.

[Paris Martineau]

Yes, yes. And it's amazing. I mean, I'm amazed on it at the level of. Of showmanship. He's. He's nuclear Steve Jobs.

[Leo Laporte]

See, I thought it was kind of awkward, to be honest with you.

[Paris Martineau]

That's part of the charm.

[Leo Laporte]

Because he refuses to prepare. There is no script, really.

[Paris Martineau]

He says there's no script, but I think he's prepared.

[Leo Laporte]

There's no prompter.

[Paris Martineau]

How do you know what slides you're gonna have?

[Leo Laporte]

Well, he just has to press the button and then he looks back and says, oh, now we're gonna talk about. So I guess that's a script. Yeah, but he is. Basically, he said at the beginning, we're doing this without a net. I don't have. We don't rehearse. It's the exact opposite, the antithesis. Now, speaking as somebody who doesn't rehearse, I think that's just laziness.

[Paris Martineau]

Yeah, that's like he rehearses.

[Jeff Jarvis]

It's definitely laziness because it takes a lot of PR people and a lot of meetings to get these sort of events together.

[Paris Martineau]

Well, at the end, when something didn't work, he said, you know, I need a human. Yeah, human. Which was a good line.

[Leo Laporte]

So gtc, which is Nvidia's, it used to be their GPU technology conference. Now it's just gtc, but is where Nvidia, who is, you know, without doubt a leader in the hardware that's used to create AI these days and software to some degree. Yeah, absolutely. So his announcements here are very, very important. He showed, for instance, new chips, new ways. This is a. A motherboard that is this size of the Blackwell processor into a much smaller 1U rack mountable motherboard.

[Paris Martineau]

Let me ask you, Jason. I said this to Jason today, even though I don't know what this does. I couldn't use it. It's too expensive for me. The geek in me says, I think I want one of those.

[Leo Laporte]

Well, in fact, he did kind of talk about the idea that we might be able to have home AI servers in the near future. They are selling a $3,000 home AI server right now, or will be soon. I don't know if you need a home AI server, but maybe you do.

[Paris Martineau]

Well, you, you, you destroy humanity, I guess.

[Leo Laporte]

Well, no, that won't. That, that'll never be a. Big enough. That's the kind of a. That's an AI tool.

[Paris Martineau]

Flat.

[Jeff Jarvis]

Watch those together and Jeff's gonna have. Have humanity destroyed in a weekend.

[Leo Laporte]

That's an AI tool. But what. But what they wong is showing is the ability to cluster thousands or hundreds.

[Paris Martineau]

What struck me so much about this was that when I watched with Micah, what got me was the scale of it. And he and I talked about that a lot at the time. Whoa, this goes on then. The last one I watched was all about the digital twins and how there's systems to check out every possible alternative future. That's the matrix. We're not in this time. It was all. If you go to the very beginning of the full presentation, it's tokens, tokens, tokens, tokens, tokens. And it's not about data, it's not about information. It's about factories to make tokens, which is really about making synthetic data to run these things.

[Leo Laporte]

Yeah, they're acquiring a company that creates synthesis synthetic data. I think they're all tied together though, Jeff, because.

[Paris Martineau]

Oh yeah, you know, they absolutely are.

[Leo Laporte]

No, tokens are a creation of these machines.

[Paris Martineau]

It's the emphasis. But, but rather than saying. He said that a computer is no longer a retriever of files, it is a generator of tokens.

[Leo Laporte]

Well, so, so what he's talking about is that these things are transformers. And the larger the, the size of data that you can handle, the more data you can handle at one time. The.

[Paris Martineau]

But it's more than just the data you handle. It's the, it's those relationships that you know of that you use of bytes of information are used to generate one token. The new chip, he said, will do 12 billion tokens per second.

[Leo Laporte]

Well, that's right.

[Paris Martineau]

In a 100 megawatts factory, as he kept calling it a factory. And so it was, it was just really interesting to me. And I don't know whether this is good in the sense that it seems to free from the bounds of the data we now have in training. On the other hand, where's the reality to it if it's all synthetic, if it's all made up, how do you get to ground truth to know that it's. It knows what real is? And, and he talked about that too. One of the interesting things was he, because he talked a lot about robotics and he called that embodied AI. And he said that you need verifiable rewards to train the robotics and that means that it's the laws of physics. And he said, we need a physics engine. So he announced a partnership with Disney to work on a physics engine to help train it. We've talked about this. The ball goes over the table. Yes, it still exists. The egg drops, if you can afford to buy one, and it breaks. What happens in real life and how to understand that. So I just, I just find him fascinating and I find his show fascinating. I don't understand a lot of it. A lot of it is high math.

[Jeff Jarvis]

Wait, what is Disney going to do to. To teach the AI about the egg?

[Paris Martineau]

I think kind of how. Yeah, it's Pixar. Yeah, I think so.

[Jeff Jarvis]

What?

[Leo Laporte]

So we've always had physics engines. Every video game has to have physics engines. Has appropriate arc parabola and ends up at the right place at the right speed. And I presume that's what we're talking about.

[Paris Martineau]

We're talking about.

[Leo Laporte]

Yeah, it's just better physics engines. So really, to me, what's going to, what's going on here? I mean, obviously he's building bigger and bigger machines, more and more information. We have seen a crisis of data. Once you eat the whole Internet, what do you do? And then there's the issue of copyright. And, and we have a number of stories, Hollywood musicians, all the creators are saying, stop reading our stuff. I disagree with them, I think they're wrong. But that's become a problem, is protecting copyright. So synthetic data is a way of getting more data, training new models based on generated data, as opposed to taking data from other people. But what's interesting is it's, I believe, is just standing on the shoulders of, of the data you already stole. So it's kind of a snake eating its own tail.

[Paris Martineau]

Not with robotics, though.

[Leo Laporte]

Well, that's okay. So that's, to me, the most interesting thing. And he brought a little robot out, didn't say anything.

[Paris Martineau]

That's a gimmick. But the interesting thing is how they train cars.

[Leo Laporte]

This is going to be the year of robotics, because that's where you get the new data. But the machine in the box does.

[Paris Martineau]

That's the physics.

[Leo Laporte]

It's looking at language, it's looking at writing. But what it doesn't know very well is what happens when the egg falls off the table. That's why you give me, you take these machines, you bring them out into the real world and you give them. And that's what Gemini, what Google is showing with this new Gemini platform, this robotics platform. We talked about this last week. This is going to be the year you see a lot of robotics. And he agreed, right? He talked a lot about that.

[Paris Martineau]

Oh yeah, he did, he did, absolutely. He says it's going to be the next trillion dollar business or 100 trillion dollar business or whatever. It was just a little minor point. He named the earlier chip after Grace Hopper.

[Leo Laporte]

It's always women.

[Paris Martineau]

Yeah, well, no, it's not going to be. David Blackwell was the next one.

[Leo Laporte]

Oh, okay.

[Paris Martineau]

And he was an American statistician and mathematician. The next is Vera Rubin, who discovered dark matter and her grandchildren were in the audience, which is kind of cool. And then he dropped real quickly that the next generation is going to be named after Richard Feynman.

[Leo Laporte]

Oh, good.

[Paris Martineau]

That cool.

[Leo Laporte]

Yeah.

[Paris Martineau]

He said there are 30 million software engineers globally. They're all going to have coding aids from all this. I think that's probably true.

[Leo Laporte]

Marcus. Gary Marcus, who was on our show a couple of weeks ago, said the synthetic data is fine when you're talking about physics and math coding, because you can generate more data just using the rules of physics and math and coding. It's a little bit more difficult on general purpose reasoning and open ended domains, as one would imagine. I think robots are a very promising area. Although again, I don't think Anthony's wrong. A very scary area because now they're out in the world. And while I'm not too worried about an LLM hurting me robot. Have you seen the new Boston Dynamics robots?

[Jeff Jarvis]

You raised something you showed this last week.

[Paris Martineau]

Yeah, you raised something last week. Paris, during that. Is that. Why are they humanoid and why.

[Leo Laporte]

Well, because of the environments that they need to. Depends what they're going to do. But if you are, for instance, a bomb defusing robot, you might need to go upstairs, open doors, move around in spaces that are too small.

[Jeff Jarvis]

Modeling them after the human body in the vast majority of cases is not the most efficient.

[Leo Laporte]

Well, that's why there's dogs. That's why the robot that Jensen Huang brought out on stage was just a little thing on a box on wheels.

[Jeff Jarvis]

Honestly, we should make more robots.

[Benito]

The robot you need to be.

[Leo Laporte]

Did I actually show this video?

[Paris Martineau]

Benito's trying to get it.

[Leo Laporte]

Yeah, I know. Did I show this video?

[Jeff Jarvis]

You did. And we talked about it at length.

[Leo Laporte]

Did I really? Did I show him two doing somersaults and running around?

[Jeff Jarvis]

He was riding a bike. Maybe. Is that this One or no, actually you didn't show the crawler.

[Paris Martineau]

They all look alike.

[Jeff Jarvis]

That's kind of fun. They kind of do.

[Leo Laporte]

I think I showed this one. This is a newer one. Benito, what were you saying?

[Benito]

I was just saying the robots you really need to be scared of are the drones, the flying ones.

[Leo Laporte]

There's all kinds of robots you might want to be scared of. So yeah, I think that's whole point of robots is they can have all sorts of form factors. A human form factor is useful for exploring spaces designed for humans, that's all. But drones are great.

[Jeff Jarvis]

So are scared of a robot when it can do a really impressive lip sync like two drag queens on.

[Leo Laporte]

How about that? How about that? That's better. That's better break dancing than a dragon ball.

[Jeff Jarvis]

Do a death drop, then a somersault over a competitor in an improvised move. That will be. That'll be the moment.

[Benito]

I'm scared also of like individuality. Like, are they actually individuals or are they one thing?

[Paris Martineau]

They're a network, aren't they?

[Leo Laporte]

Yeah.

[Benito]

So like, what do you mean individuals?

[Leo Laporte]

I don't understand.

[Benito]

Are the robots, are they individual, like, for lack of a better word, entities, like people? Or is that. Or do we count them as all one thing, like a hive?

[Jeff Jarvis]

I think it depends how they're being controlled. Yeah, you control how they've been designed. I mean, if it's like a one to one sort of control thing where a robot, a person has to be like robots going to take a step now and then do a somersault, then that's an individual robot. But if you have like 10 of them moving en masse, you'd probably refer to them as a.

[Leo Laporte]

You might ask the same question about an army brigade. Are they individuals or are they, each of them individuals?

[Benito]

Don't we. We consider each of the people people. Individual people.

[Leo Laporte]

I think we train.

[Jeff Jarvis]

Are you saying that you don't believe the military that are there are armed forces is made up of individuals?

[Leo Laporte]

I think we train them extensively so that they act as a unit, not as individuals. The one thing you don't want like.

[Benito]

That, but you know, you don't want.

[Leo Laporte]

Soldiers to act as individuals because individuals protect their individuals, though they are.

[Jeff Jarvis]

Yeah, Leo hates the troops. That's what we're learning here.

[Leo Laporte]

No, I'm saying that I'm joking.

[Paris Martineau]

All right, well, here's. Here's where it's coming closest to you. Here's where it affects your life.

[Leo Laporte]

Life.

[Paris Martineau]

Here's where AI is really going to matter and Nvidia is going to Come into your life next time you go order at Taco Bell.

[Leo Laporte]

Is this a story? What?

[Paris Martineau]

Yes, it is. Yes, it is. Brands. I've already Yum brands. There have been various efforts. McDonald's did it. Wendy's did it. Now Yum Brands is those were failures. Now they're partnering with Nvidia to do.

[Jeff Jarvis]

AI ordering to scale the fast food company's existing bite by yum platform.

[Paris Martineau]

Great brand, huh? Great brand.

[Jeff Jarvis]

Spelled B Y T E. Of course. Biome. What are they actually doing though? Park said the companies plan to roll out AI solutions to about 500 restaurants across the four Yum brands, including Habit Burger and Grill. In addition to Cash, which is very.

[Paris Martineau]

Good by the way.

[Jeff Jarvis]

Have a Bird and Pizza Hut.

[Paris Martineau]

Interesting.

[Jeff Jarvis]

Wait, no. But what are they going to do?

[Benito]

Last time I drove through Burger King, I talked to an AI.

[Paris Martineau]

You did? Yep.

[Jeff Jarvis]

You did.

[Paris Martineau]

Did you try to screw with its head?

[Benito]

No, I didn't. I mean I wanted to see.

[Paris Martineau]

You wanted your bur.

[Benito]

I wanted. I wanted to see what actually worked properly. So I just did the.

[Paris Martineau]

I was supposed.

[Jeff Jarvis]

Wait, have you guys been to one of the new KFC drive thru like locations where they have you pull up in like a preload area and. Well, when I was there, some teens with like computers on their chest took my. And then you got into a special lane and then you went through like a whole thing. It was a very streamlined experience.

[Leo Laporte]

You are such a New Yorker. They've been doing that at in and out for 10 years. Yeah.

[Jeff Jarvis]

Really?

[Leo Laporte]

You have to go to the suburbs because there's lots of land people with.

[Jeff Jarvis]

People outside standing on the street 24.

[Benito]

Hours, it gets real long.

[Paris Martineau]

You go to the in and out by SFO and they're always, always jammed and they're always there.

[Leo Laporte]

Yeah, they have to be. Otherwise you'd be sitting in the line for half an hour without anybody to talk to.

[Paris Martineau]

And you, the neighbors will be pissed off because the traffic goes up back way.

[Leo Laporte]

I don't think it's up.

[Paris Martineau]

Okay, I want to hear do you like it or not burgers, by the way. I never heard ruling on that.

[Jeff Jarvis]

I've never.

[Paris Martineau]

You know where you went out burger.

[Benito]

Yeah, of course. I love in and out burgers.

[Leo Laporte]

Who doesn't love in and out burgers?

[Paris Martineau]

They're okay. Oh, oh, oh, Leo, Leo, I've got to tell you this. Oh, well, when you're in New York for Hank's opening.

[Leo Laporte]

Yes.

[Paris Martineau]

You need to Paris. I'm curious whether you've been here. You need to desert him for just one meal and Go to American Burger.

[Leo Laporte]

Is this the one where you go through the hotel lobby to get to the little kiosk in the back?

[Paris Martineau]

That's all. No, American Burgers.

[Jeff Jarvis]

I haven't had a burger from American Burger there.

[Paris Martineau]

Hamburger America.

[Leo Laporte]

Speaking of stories we've done. Definitely talk. Talked about American Burger before.

[Paris Martineau]

No, Hamburger America. No, I don't.

[Leo Laporte]

Hamburger America. Okay.

[Paris Martineau]

Hamburger America.

[Jeff Jarvis]

They.

[Paris Martineau]

I'm telling you, they do the onion burger. It's a smash burger.

[Anthony Aguirre]

Right?

[Paris Martineau]

You.

[Leo Laporte]

I do like smash burgers because my.

[Paris Martineau]

Art with the onions are just in it, like, crazy.

[Leo Laporte]

Is this the one on McDougal Street?

[Paris Martineau]

Yes.

[Leo Laporte]

Okay.

[Paris Martineau]

Right near where Hank's gonna be.

[Leo Laporte]

Yeah.

[Paris Martineau]

And it is unbelievable. You've been here.

[Leo Laporte]

It was. It was founded by a renowned burger scholar.

[Paris Martineau]

Yes, yes.

[Jeff Jarvis]

God, I wish. I wish I could be.

[Paris Martineau]

It is so good. It is killer.

[Leo Laporte]

Really? Oh, really?

[Paris Martineau]

Scroll down that site and look at the onion burgers there.

[Leo Laporte]

This is. This is. There he is. There's the renowned burger scholar himself. He's got sideburns to write. Oh, look. See, a smash burger starts as a ball, and then they squish it.

[Jeff Jarvis]

But the onion does look like a burger Scholar.

[Paris Martineau]

It does.

[Leo Laporte]

You start to render the fat. Then you salt it, and the salt mix in with the rendered beef fat. Then you add the onion. That's a lot of onions.

[Paris Martineau]

Oh, it is.

[Jeff Jarvis]

That's the right amount of onions.

[Paris Martineau]

They really caramelize the heck out of it.

[Leo Laporte]

Yeah, the onions are trapped in the beef steam. You put the cheese on the bun. Now you have the six sirens. Well, I should let the guy speak for himself. Making me hungry.

[Jeff Jarvis]

I'm making myself hungry.

[Paris Martineau]

That's George recreating classic, classic American burgers and as authentically as possible.

[Leo Laporte]

Huh? The best burger scholar, George.

[Paris Martineau]

He should have mozzarella, of course, but he doesn't.

[Leo Laporte]

I do think the smash burger is a. Is a significant improvement.

[Paris Martineau]

Oh, yeah. There's a place in New Jersey, it is white manna that goes way back to the. To the, like, 30s. And it's amazing. It's. It's. It's that old diner kind of thing with white tiles and stuff. Oh, it's killer. Killer.

[Leo Laporte]

Google is officially replacing the Google voice assistant with Gemini. This is something Apple could not do. Apple tried.

[Paris Martineau]

Standards.

[Leo Laporte]

Maybe they have higher standards. I don't think that's the problem. You know, they hired John Jannandrea away from Google, who was a highly respected AI scientist, and they've been working on this. Apple, just last week or maybe two. Two weeks ago, announced. Yeah, it's not Going so well, we probably won't have a smart Siri for another year. Google says. Hold my beer. Here's a video of Google replacing.

[Jeff Jarvis]

It says, you read that title correctly.

[Leo Laporte]

The time has finally come. The Google Assistant on your smartphone. Google is replacing me with Gemini. Wow. Wow. So I think I saw this. You could choose your. I think I should go get my. You use a Pixel, right? Jeff, have you seen this?

[Paris Martineau]

Hello. Who am I?

[Leo Laporte]

I know.

[Paris Martineau]

Yes.

[Leo Laporte]

You're Mr. Google. You. You've. You're all in on a dying empire of Google.

[Paris Martineau]

I live la vida. Google. Yeah. How do I. Where do I go to find out? Let's search.

[Leo Laporte]

So the newest Google phones, including the Pixel, did this.

[Paris Martineau]

I have sex.

[Leo Laporte]

Google says millions of people are. But you have an older Google phone.

[Paris Martineau]

Yeah.

[Leo Laporte]

So that may be not coming to you soon. I don't know. So later this month, they say, we believe an assistance should be personal to you and aware of the world around you. See, I don't know why you think this is about to happen. Happen every. It's happening all around us and it.

[Paris Martineau]

Should be able to.

[Leo Laporte]

Oh, this is agentic. It should be able to interact with the apps and services you already use, and it should make you more productive, more creative, and a bit more curious. It's about time something made you a bit more curious.

[Jeff Jarvis]

What are you talking about?

[Leo Laporte]

I'm reading the press release.

[Jeff Jarvis]

How will this make me more curious? How are we even remotely close to a future where AI is going to kill us all because it's so super powerful and autonomous?

[Paris Martineau]

Yes.

[Jeff Jarvis]

Come on.

[Leo Laporte]

You're going to be sorry. Sometime. I'm going to be. Jeff and I are going to be long gone, but you're going to be sorry. You're the one who's going to have to suffer with all this.

[Paris Martineau]

Who was right?

[Leo Laporte]

Yeah. You're going to have to arbitrate this. An exec overseeing Siri told staff the delays were ugly, made worse by Apple publicly promoting the new features. And it's unclear when they were shipped. This is Mark Gurman and the Bloomberg rumor guru. Ugly and embarrassing. Jesus, that's bad. There's a history for that, though. Steve Jobs famously pulled the Mobile Me team into the auditorium and screamed at them, saying, you're hurting the company. He said to them, what is mobile me supposed to do? And they said, it's the cloud. Well, why doesn't effing it do it? Why doesn't it effing do it? During the All Hands gathering, Robby Walker, who serves as a senior director at Apple during an all hands gathering, suggested employees on his team may be feeling angry, disappointed, burned out and embarrassed after the features were postponed. They shouldn't feel embarrassed, he says. But he said, you developed incredibly impressive features and someday we will deliver an industry leading virtual assistance one day, one day soon.

[Paris Martineau]

What I would love to know is the spec, which of course they never released, what was the spec? What was their expectation? What do they think it was going to do looking at all the competitors, what was the specific historically?

[Leo Laporte]

And this is I think the same problem Amazon's having. It's hard to take. I think we talked about this. A voice assistant that is really there to play music, buy stuff on Amazon, set timers and to turn it into a chatgpt style AI chatbot. They're very different tracks. You can't just flip a switch and make it do that. And I think that both Amazon and Apple are having that difficulty. Anyway, here's a thing that I don't know if people are happy about, but Google's a Gemini AI is really good at watermark removal. According to Umar Shakir, writing in the Verge, it can cleanly replace the Getty Images watermark with an edited with AI watermark. If you should so choose.

[Benito]

I do this for the show every week because of the intelligent machines watermark in the top left. I get rid of that. Yep.

[Paris Martineau]

Really?

[Jeff Jarvis]

Really.

[Leo Laporte]

Why is there a watermark on our.

[Benito]

It's not a watermark.

[Jeff Jarvis]

It's replacing. It's replacing the watermark with my books.

[Benito]

The little icon on the top left, the intelligence icon. I removed this thing.

[Leo Laporte]

You have to replace that?

[Benito]

Yeah, I have to just get rid of it for the thumbnail. And I do that every week with.

[Leo Laporte]

And do you.

[Anthony Aguirre]

What do you.

[Jeff Jarvis]

What AI tool?

[Benito]

No, there's a thing in Photoshop, you just select it and say fill and it.

[Paris Martineau]

Okay, now you've got to trick it, Paris, since it's always you, you've got to put up something that is not expected, the first four letters of a word and see what it makes up.

[Jeff Jarvis]

Yeah, I do. I was imagining a large cowboy hat, but that could be. That'd probably be fun.

[Benito]

And it really does an amazing job. It like, like it really looks like it was part of the picture. That's one of those tools that makes me a believer in that.

[Leo Laporte]

If you think about it, it's a.

[Paris Martineau]

Long step in destroying mankind, but yes, it's impressive.

[Leo Laporte]

It's exactly what an LLM does, which is it's looking. There's now an area and it has to predict what the Next pixel would be exactly.

[Benito]

This is exactly what it's good at. Exactly what it's good at.

[Anthony Aguirre]

Yeah.

[Leo Laporte]

There is a long way from that to a world destroying robot. But.

[Paris Martineau]

But it's good.

[Leo Laporte]

Good.

[Paris Martineau]

It's really good.

[Jeff Jarvis]

Oh, it's just days away.

[Leo Laporte]

You're gonna be sorry when this show.

[Jeff Jarvis]

Is replaced by the show where we just threatened me.

[Leo Laporte]

You're so sorry.

[Jeff Jarvis]

You're gonna be sleeping with the Hollywood.

[Leo Laporte]

Hollywood is very worried about the AI industry's push to change copyright law. And when I say Hollywood, I mean Chris Rock, Aubrey Plaza and Paul McCartney among hundreds of others who sign an open letter to the government. Good luck with that. Challenging OpenAI and Google. They sent it to the White House Office of Science and Technology. I don't know if there's anybody there to receive it. 400 members of the entertainment community expressed concerns about the Wish lists of AI companies for the U.S. aI action plan. And we have an AI action plan.

[Paris Martineau]

Arnold Schwarzenegger, an AI action plan.

[Leo Laporte]

You need auction those recommendations that let a warrant could severely damage the entertainment community.

[Paris Martineau]

I love cueing that part of LEO. It's just the greatest.

[Leo Laporte]

Which supports more than 2.3 million jobs whose wages are currently $229 billion a year. So there. The group urged the Trump administration not to sacrifice America's leadership in the world of entertainment and the race to dominate AI. Where do you stand on this? I mean this is what Kathy Gellis calls the right to read. These actors, directors and musicians don't mind it if I go to their movie, watch the movie and then tell a friend what it was about. Or do they? Maybe they want to also legislate against spoilers. They probably would if they could.

[Jeff Jarvis]

It's not that simple. It is. I mean we've had this discussion a million times in the show. But part of what the problem is here is should a for profit entity be able to scoop up large amounts of content that typically require being like each one of the each movie if you wanted to purchase it, you'd actually have to spend money to buy a disc or spend money to go see it. Should you be able to scoop that up for free en masse and then use it to generate profits? It I would argue no. And most of these companies and products have clauses in their terms of contract saying you even if you buy a DVD of Twin Peaks like I have many times you can't. I can't. Then that doesn't mean that I can then take all of Twin Peaks and sell it part of it and use part of it in my own things. But why is it when an AI.

[Leo Laporte]

Does it, that they're not selling Twin Peaks, they're just watching.

[Paris Martineau]

They're learning how mountains look like.

[Leo Laporte]

Yeah, they're learning.

[Paris Martineau]

They're learning to speak. It's another matter if they wrote a script and it was Twin Peaks. Yes, it's another.

[Jeff Jarvis]

But it could do that.

[Paris Martineau]

Stole it. Well, but then it should. Then I would agree. There's three cases here. One is training, and to me, that is fair use and transformative. If that's all it really does, it's teaching you to speak or to draw. Two is acquisition. If you steal it, then that's an issue. But that's an issue in all law existed before. And three is quoting. If you quote it at length without licensing, that's wrong. I think we have to separate those things.

[Jeff Jarvis]

I think part of the issue, though, is these systems and tools are not learning. They're not reading, they're not watching a movie. They are their code bases. It is a dataset that is being improved in some ways and then having prediction capabilities based on that content. And where is it getting the, like, output that it is getting it is getting it from the things that are put into it. It's not coming up with new words and new thoughts. It's ingesting material and then making predictions based off of that, based on.

[Paris Martineau]

Based on a trillion tokens, as my. My hero Jensen said. So. So it's. It's the applicability of who got what. Let me ask a different question, Paris. So let's go with the presumption that we're going to all have AI everywhere. It's going to be in charge of the world. It's not necessarily bad, but we're gonna have.

[Jeff Jarvis]

I would disagree with that presumption, but, you know.

[Paris Martineau]

Well, I did earlier, too. I'm just trying to, you know, go with the punchline, go with the joke. So if we, if we presume that society is going to use AI like crazy, do we want smarter AIs or stupider AIs? Is there. Is there a question of enabling them to be better?

[Jeff Jarvis]

Because I'll give you. I'll give you another question in response to your question. We're back in the day of the horse, and Bucky, this dear old guy comes up, says, I've got this newfangled invention called the car. Imagine that someday in the future, soon, I can tell you with absolute certainty, everybody's going to be wanting to drive one of these, using it to get everywhere. Do you want your car to be able to get you places quickly, much like the horse, but even faster. Well, then we shouldn't ever even think about these things like seat belts or windshields. That could just, you know, slow down the innovation. That could just slow down the innovation of our here car product. So we should really focus on the stuff that makes it better.

[Leo Laporte]

If it didn't watch Twin Peaks, could I do this? Can you explain what the hell's going on in Twin Peaks? Don't.

[Jeff Jarvis]

I haven't seen the last season.

[Paris Martineau]

You can't do that.

[Jeff Jarvis]

Don't let it.

[Paris Martineau]

Mute yourself.

[Anthony Aguirre]

Television series created by David lynch and Mark Frost.

[Leo Laporte]

And it's known for its complex, surreal.

[Anthony Aguirre]

And often unsettling narrative.

[Leo Laporte]

It's like your friend here's a breakdown of what's going on.

[Anthony Aguirre]

The basic premise. The series begins with the murder of.

[Leo Laporte]

Laura Palmer, pop student in the seemingly idyllic town. Now, if it hadn't watched Take it out, would it be able to give you this valuable information?

[Paris Martineau]

Well, that it got from lots and lots and lots of reviews.

[Leo Laporte]

Hey, Palmer, do you know.

[Jeff Jarvis]

Hey, if you haven't listened to watch Twin Peaks, pause the podcast right now.

[Anthony Aguirre]

Palmer is complex as it involves both a human perpetrator and a supernatural entity.

[Leo Laporte]

What? I didn't know that.

[Jeff Jarvis]

Yeah, there's a lot of stuff I didn't know when I started Twin Peaks, but I was really surprised by, and it was delight. If you. You somehow, for the last 40, 30 years, haven't watched Twin Peaks, don't let Leo ruin it for you.

[Leo Laporte]

No, but look it. I love Paul McCartney, Cynthia Erivo, Clay Blanchett. These are all signatories to this letter. Phoebe Waller Bridge, Bette Midler, Cate Blanchett, Paul Simon, Ben Stiller, Aubrey Plaza, Ron Howard, Taika Waititi, Ayo Edebiri, Joseph Gordon Levitt, Janelle Monae.

[Paris Martineau]

All right.

[Leo Laporte]

Rian Johnson, Paul Giamatti, Maggie Gyllenhaal, Alfonso Cuaron. These are the. These are the cultural icons of our generation. Olivia Wilde, Judd Apatow, Chris Rock, Mark Ruffalo, all of them. But then if I said but, okay, but do you not want me to go see the movie either? They would say, no, please go see our movie. Look, I understand this is an existential threat to them. I completely understand and I feel for them. But should we let the buggy whip makers prevent us from getting the Model T?

[Jeff Jarvis]

I don't think that's what's. I think people are saying the Model T shouldn't, you know, follow the rules. The fall Model T should follow the rules that Everybody else has to follow.

[Paris Martineau]

Right, well, okay, let's stay there. Journalists read each other and probably journalists don't have a.

[Jeff Jarvis]

Journalists do not have a photographic memory of every single thing that they have read.

[Leo Laporte]

That's true. The AI will not.

[Jeff Jarvis]

And then they are not immediately. That photographic memory is not immediately searchable and interactable by any Joe Schmo in the world. For right now, free, but soon to be. You know, it's a paid product.

[Leo Laporte]

I think. Look, you both are right, but the good news is when the super intelligence comes, we won't have to worry about any of this because we'll be dead. We'll be. No, we might not be dead. We might just be batteries. You know, they're going to need somebody to feed the coal into the burners or something. Whatever it is, the nuclear rods into the reactors. All right, let's take a little break.

[Benito]

Sorry, sorry.

[Leo Laporte]

Yeah, go ahead, take your break, Benito, please.

[Benito]

I think what it really comes down to, it's like, not necessarily that the computer's reading is that they're charging us back the stuff that we put into it. That's what I disagree with. Like, why are you going to charge me for this? I put the data.

[Leo Laporte]

How did you learn to talk, Benito?

[Benito]

We all put the data.

[Leo Laporte]

How did you learn to talk? Didn't you listen to a bunch of people? Did you pay them back, by the way, for the tokens properly? Okay, that's good. You went. You went to a better school. I mean, people. This is what. This is how it happens. What they complain is not that it's happening to people, it's happening to machines is what people don't like, the machines are doing it. But believe me, when the machines take over, you're gonna be sorry you were so negative about the machines. I can just.

[Jeff Jarvis]

It's not even that machines are doing it. It is that machines that are the product of a. Of incredibly large and well funded corporations are doing it. It is that.

[Leo Laporte]

I wouldn't want to.

[Jeff Jarvis]

It's that all of our art and creative output and anything in humanity, even stuff that people are wanting to keep either private or at least behind a paywall, is being fed into someone else's product and they are making money off.

[Benito]

We have all contributed to it. We have all contributed to this thing.

[Paris Martineau]

It's called culture.

[Benito]

Exactly.

[Paris Martineau]

Our culture.

[Benito]

So why are we getting charged for it then? That should come back right to us, right?

[Paris Martineau]

Well, every. Every novelist learns from earlier novelists. Every filmmaker learns from other filmmakers. How many filmmakers owe Alfred Henry Hitchcock money like that's?

[Benito]

I think that's a false equivalence because.

[Paris Martineau]

Why they learn from him. They use his same techniques.

[Benito]

It's a difference of scale and it's a difference of kind. Like it's a difference of both things.

[Paris Martineau]

Let's try.

[Leo Laporte]

No, we can disagree.

[Jeff Jarvis]

I think reasonable disagreement has a very valid point.

[Leo Laporte]

Yeah.

[Jeff Jarvis]

Is that, you know, a random filmmaker or a very important filmmaker, whatever. The scale is not the same as a company that is raising trillion doll millions of dollars on the promise of we will steal everybody else's stuff.

[Paris Martineau]

So let's say the filmmaker is Disney. Oh, no. It's a huge corporation.

[Leo Laporte]

Well, I'm really. Yeah. I'm a little frosted that Disney stole all the Grimm Brothers material and then copyrighted it.

[Paris Martineau]

Yeah, yeah.

[Jeff Jarvis]

Disney famously is. People get mad at Disney and sue Disney when Disney does this stuff. So I would agree. Disney's a great example of this. We should treat the AI companies like we've treated Disney. Disney.

[Paris Martineau]

Well, that's to say, pass laws that do just what they want to do is where that goes.

[Leo Laporte]

All right, let's. We do have to take a break. You're watching a very good argument. And you know, I wish there were an answer to all of this. And I don't think there is. That's the thing. And there's. Reasonable people can disagree. And I don't know. I don't know.

[Paris Martineau]

That's what we're here to do every week.

[Leo Laporte]

That's what the answer is every week. Jeff Jarvis is here. He is. I didn't get a chance to do this earlier, so I will do it now. The emeritus professor of journalistic innovation at the Craig. This is the rag Newmark Graduate School of Journalism at the City University of New York. He is now at Montclair State University and SUNY Stony Brook. Great to have you, Jeff. And of course, Paris Martineau, who is filing a big story. You were working hard this week at the Information.

[Jeff Jarvis]

I was.

[Leo Laporte]

Anything you could tell us story coming.

[Jeff Jarvis]

This weekend allegedly about section 230. Keep it.

[Paris Martineau]

Oh.

[Leo Laporte]

Oh, good.

[Paris Martineau]

Is. I got one question for you. Is Mike Masnick going to be proud of you?

[Jeff Jarvis]

Yeah.

[Paris Martineau]

You know about Mike's famous page?

[Jeff Jarvis]

Yes, of course.

[Leo Laporte]

Yes.

[Jeff Jarvis]

I've listened to his podcast, which is recently. Okay.

[Paris Martineau]

I'm just making sure Mike. So if Mike's. Mike is God when it comes to.

[Leo Laporte]

He's the defender of the 23 words.

[Paris Martineau]

That is King Arthur. 236.

[Leo Laporte]

Yes. All right, good. I think this is good. I'm glad. I can't wait to read it. That's exciting. That's really good. This is a lot of. You know, it's under assault right now, obviously.

[Jeff Jarvis]

Yes.

[Leo Laporte]

Oh yes, it has been since, for the last decade. This episode of Intelligent Machines brought to you by our friends at Zscaler, the leader in cloud security. So the situation we're in right now, not so hot. Enterprises have been spending billions of dollars on perimeter defenses, firewalls, right? And then VPNs so that you can get into, work through, through the firewall. Has this fixed everything? No. Breaches continue to rise. There was an 18% year over year increase last year in ransomware attacks. You think it's going to be less? No, it's going to be worse this year. Last year, a record payout, $75 million in ransomware ransoms last year. And that is going to go up, of course, this year. See, the problem is traditional security tools actually make you less secure. They expand your attack surface, giving you public facing IP addresses that bad actors can go, yeah, there they are, and hang their hat on it. And now they're doing it more easily and faster than ever with AI tools. Plus, once the bad guys get in, most security tools just assume, well, they must be an employee. So they let them have at it. They can access all the software, they can access all the data, they can move laterally throughout your entire network. Find the stuff you'd be most embarrassed about. Exfiltrated encrypted packets through your firewall, which is struggling to inspect those encrypted packets. And you've got a problem. The fact is, hackers are exploiting our traditional security infrastructure and they're doing it faster than ever with AI. It's time to rethink security. You can't let these guys win. Well, that's why you want to check out Zscaler. Zero Trust plus AI puts AI on your side. First of all, it stops attackers by hiding your attack. Surface apps are invisible IP addresses. Invisible bad guys can't attack what they can't see. It also eliminates lateral movement because users can only connect to specific apps, apps they've been approved to connect to, not the entire network. And it continuously verifies every request based on identity and context. Plus, it uses AI to make your life better. It simplifies your security management using AI powered automation. And of course, course, Zscaler uses AI. They analyze over half a trillion transactions every single day. And they use AI to find the real threats because the mass of those transactions are legit. The needles in the haystack. The real threats and that's where AI comes in really handy. So bottom line, hackers, they can't attack what they can't see. It's time to protect your organization. Hide those IP addresses. Hide those apps from the bad guys. With Zscaler Zero Trust Plus AI learn more at zscaler.com Security Put AI to work for you zscaler.com Security we thank them so much for their support of intelligent machines. There was an interview with. It's not just a movie stars video game voice actors are also losing their jobs because of course, in fact it makes the games better. Let's face it, a video game actor might have to come in and record 50,000 lines. So there's a variety in your. In your. They talk, don't they? In your. Baldur's. Baldur's Gate 3. They talk, yeah.

[Jeff Jarvis]

And the actors, the voice actors were a really integral part of character development. They did motion cap and they. I think it's one of the most captivating performances I've seen in a video game, at least in recent memory. Because the voice actors were so integrally involved. All the different line reads and all the different changes which wouldn't happen if you just used AI to crank out any new change.

[Leo Laporte]

Well, Sony thinks so. Ashley Burch is the award winning voice and performance actor behind Alloy in Horizon's Zero Dawn.

[Jeff Jarvis]

A video game, may I add, about AI super intelligence being essentially destroying the world because it is used for war machines. Then they become autonomous and take over the world and rid it of human life.

[Paris Martineau]

Meta on meta.

[Leo Laporte]

So there was a video, it's now taken down, of a Sony experiment. A leaked video that Shaw saw Alloy voiced in performed by AI technology. Ashley saw it and was not happy. She said, well, I'll let her speak in her own words.

[Paris Martineau]

I wanted to talk about AI Aloy.

[Leo Laporte]

Oh, it's aloy. I mispronounced it. Oh, she got quiet.

[Jeff Jarvis]

It's a long trend. It's a long video.

[Leo Laporte]

Okay, so you get the idea. She says having seen the demo, she's worried and not just about her own career. She's worried about the art. Art form. Yeah, there's a whole transcript.

[Jeff Jarvis]

Yeah. I mean, I think in this case they said it was just an internal test. They have no plans of replacing her. Blah, blah. But I think this is still a really important issue. I mean, members of Strike related to this.

[Leo Laporte]

If you're right and Baldur's Gate 3 is, or 4 is not as good because of AI voices, it will hurt them commercially, won't it?

[Jeff Jarvis]

I mean, in the long run, a bunch of people are still going to buy Baldur's Gate 4, even though it's not going to be produced by Larian Studios, which made Baldur's Gate 3 and the other ones. So it's going to be bad already. Larian made the third. Sorry.

[Leo Laporte]

Okay. I mean, that's the thing. If there. If the AI is not as good, it's not going to work. If, on the other hand, an AI is as good as a real actor.

[Paris Martineau]

Well, this is you. Last week, thought that OpenAI's story was good. We made fun of you.

[Leo Laporte]

You laughed at me. But you know what? I'm not alone.

[Paris Martineau]

No, you're not.

[Leo Laporte]

There's a Guardian, a writer, a real, live, actual writer.

[Paris Martineau]

For Christ's sakes.

[Leo Laporte]

She's a novelist. Jeanette Winterston, writing a piece in the Guardian, said that metafictional short story that Sam Altman presented and I read a little bit of last week is beautiful and moving. So there. She says she doesn't like AI. The alternative to it she prefers is alternative intelligence, not artificial intelligence. Because in all the fear and anger foaming around AI just now, its capacity to be other is what the human race needs, she says. I should have mentioned this to Anthony Aguirre. Our thinking is getting us nowhere fast except towards extinction or planet via planetary collapse or global war. We need this alternative brain to help us get past. She's not wrong. Humans have not done a good job.

[Paris Martineau]

Yeah, that's. Interesting argument.

[Leo Laporte]

There's been a lot of fuss. She writes, and rightly so, about robbing creatives of their copyright. To train AI, tech bros need to pay for what they want. They pay lawyers and lobbyists pay artists. It really is that simple. Okay, but what is not as simple is the future of human creativity as AI systems get better at being creative. H. She likes it. AI reads us. Now it's time for us to read AI, she says.

[Benito]

But it means nothing.

[Jeff Jarvis]

Yeah, well, that's.

[Leo Laporte]

You made that point, Benito. That the short story, however well written, if it's not representing a human point of view, if there's no heart beating behind it, it's just not as good.

[Benito]

The point of art is for humans to imbue meaning into it, right? Like, then if. If that. If that part doesn't happen, what's the point?

[Leo Laporte]

No, I agree. Let humans make art. Let the AIs make paperclips. Fine. I don't think AI should be. I don't. You know, the AI image generation and all of that. The music generation. Let's not. That makes. It makes wallpaper. Let's make. Let's. Humans should make the art. I agree. And. And we should have more time to make art because the AI is putting our babies to sleep.

[Jeff Jarvis]

So you agree that with their copyright concerns.

[Leo Laporte]

Yeah, I mean, I understand their copyright concerns. And you're right, I agree with you. If it's somebody, some big company making a lot of money off of it, it's the least they could do is at least pay for a ticket to the movie. Yeah.

[Benito]

I think this is the fundamental disagreement we're having is that we're not really talking about the technology. We're complaining about the company. Companies.

[Leo Laporte]

The companies. Yeah.

[Benito]

The technology is the technology. It's neutral, right?

[Jeff Jarvis]

Yeah. Largely.

[Paris Martineau]

Well, never fully.

[Leo Laporte]

I have to say, one book that I am enjoying reading is this Band Tome.

[Jeff Jarvis]

Oh, yeah.

[Leo Laporte]

So the minute I read that Meta and Mark Zuckerberg were trying to shut Sarah Wynn Williams down for her memoir of her time at Meta, I just had to buy it. It's called the Stress and Effect and it really works. I had read an excerpt already about her. You know the one where Sheryl Sandberg invites her to join her in their. In their first class airplane bed.

[Paris Martineau]

Boudoir.

[Leo Laporte]

Boudoir. And there's a lot of. It's juicy as hell. Paul Thurat's reading, he did the same thing. He said, well, if they're going to ban it, I better buy it now. Now. So it's probably sold a lot more because of that.

[Paris Martineau]

Oh, yeah. Oh, yeah.

[Leo Laporte]

So dumb. It's juicy. He doesn't like it because there's no technology in it. Well, no, it's just a. It's a. It's a. It's, you know. Sex.

[Paris Martineau]

Robot. Throat.

[Leo Laporte]

Yeah, he's a robot. I haven't, I have only read a little bit of it, but it's. Yeah, it's tea. Spilling the tea. But it just shows you if you don't like something, the worst thing you can do is sue them so that you get a lot of headlines.

[Paris Martineau]

Do you know Katie Harbath?

[Leo Laporte]

I know that name. Why do I know.

[Paris Martineau]

So she was, she was a policy person at Meta, formerly known as. As Facebook. Facebook. And she's left. I think she's at the Aspen Institute now. So she wrote a piece and she thinks that Joel Kaplan was a wonderful boss.

[Leo Laporte]

She's, she's, you know, know, she countered the narrative.

[Paris Martineau]

She counted the narrative. She said there were some things that were true and some things that were sloppy and wrong. And, you know, you gotta Go back and forth.

[Leo Laporte]

We just don't know. I mean, honestly I. I don't know if it's true or not, but I can. All I do know is when you try to suppress a work of. Yeah, non fiction or fiction, you're just gonna make people want to read.

[Paris Martineau]

The amazing thing about is you think that Joel Kaplan, who's in charge, who's now the new president of pr, basically there's. Would know better, would be more sophisticated about this.

[Leo Laporte]

Yeah, it's not very sophisticated. They got an arbiter to say that the author cannot talk about it, promote it. They couldn't really stop the publishers. They tried.

[Paris Martineau]

They can't. No, they can't.

[Jeff Jarvis]

Well, I mean I know so many people who've started reading it in the last week just because of this news cycle. Like literally as we were getting on this podcast, the Skeeball group chat was popping off, people being like, oh, I'm listening to the audiobook of careless people now. After I heard about the suit, it's such a, a classic example of the Streisand effect at work.

[Paris Martineau]

So when I left Time Inc. When I walked out of Entertainment Weekly, they were happy to see me go. They would have fired me. I quit. But I would not sign the managing editor's contract at Time Inc. Because it had an NDA.

[Leo Laporte]

Yeah, yeah, yeah.

[Paris Martineau]

And I said it's wrong for a journalist to do that. If I had signed it and if I had arranged for them to fire Little check. I bet 3 years salary, bonus and benefits.

[Leo Laporte]

You are a man of principle.

[Paris Martineau]

I know the price of principal to the petty, young and foolish that I.

[Leo Laporte]

Was three when we of course, as any sensible company does, put a non disparagement clause into our severance agreements. And usually it is tied to like you want your severance, but I think we've waived it. I think we don't.

[Jeff Jarvis]

In California that's also illegal.

[Leo Laporte]

Now is it a non disparagement?

[Paris Martineau]

I thought it was just a tied to severance. Ah, really? Oh, well, this is the reason.

[Leo Laporte]

What we were talking about is probably what they were giving Jeff, which is over and above the legal requirement.

[Paris Martineau]

Well, this was many years ago, so.

[Leo Laporte]

Yeah, they didn't have to give you three years or offer you three years.

[Benito]

I'm sure three years is ridiculous.

[Leo Laporte]

That's a lot. That's a lot. Scolded parachute and have you disparaged them? Have you? Have you? Actually that's the thing.

[Jeff Jarvis]

Well, he went into it in depth in his book magazine. I knew this story from magazine magazine.

[Paris Martineau]

Right Because I. I had. I kept the documents that I wrote about it. Now, of course, the company is basically.

[Leo Laporte]

Dead and gone, so nobody cares.

[Paris Martineau]

So nobody would have cared the other day. So when I was a Conde Nast at Advance, I said, fire, because they fire really well. Oh, they fire so well. But they wouldn't fire me.

[Leo Laporte]

Damn it.

[Paris Martineau]

I tried. God damn it.

[Leo Laporte]

So, as you know, I like Perplexity. But maybe now I should think again.

[Paris Martineau]

Yeah, this was interesting, wasn't it?

[Leo Laporte]

Yeah. The Daily Wire has announced a new advertising partner partnership with Perplexity and my favorite guy, Ben Shapiro. I don't understand. Is this an ad? What is he doing? How is this. How does.

[Jeff Jarvis]

So he's advertised. He's reading ads for Perplexity.

[Leo Laporte]

It won't utilize Shapiro's quintessential rapid fire reads. Shapiro will integrate the AI technology into his. How come I can't get some money into his regular reporting and utilize the tool to ask questions in real time, getting information straight from the Internet to supplement his commentary. He's getting paid to do what I'm doing. I'm paying them. Oh, well, he's. He's a big star, right?

[Paris Martineau]

In certain circles.

[Leo Laporte]

In certain. Certain circles. Apparent. Now, I don't know if this is true. You put this in here. Is Grok gonna merge with.

[Paris Martineau]

I found that just hard to believe. But I put it in here because it's.

[Leo Laporte]

This is from a. Who is the American?

[Paris Martineau]

I don't know. I don't know. So, yeah, this is Perplexity might be.

[Leo Laporte]

Looking to partner up with Grok. Apparently, Perplexity CEO Aravind Srinivas hinted at a potential integration of the two AI models. This is his X post. Let me read it. It says hopefully soon go up. No.

[Paris Martineau]

Or down. Whatever you call that, since you're such a bad scroller.

[Leo Laporte]

The two best. So Aravan is saying the two best AIs now for real time information. Okay, so this is a complete misunderstanding of how Perplexity works. Perplexity has multiple models.

[Paris Martineau]

That's what I was thinking.

[Leo Laporte]

You can choose from in Perplexity.

[Paris Martineau]

So it just adds another model you could choose from.

[Leo Laporte]

So I'm guessing that this is not a merger of Perplexity. And that's a misunderstanding. It would just be that Korok would be one of the engines I could choose.

[Paris Martineau]

That makes me feel better.

[Leo Laporte]

I'm sure that's the case.

[Paris Martineau]

It's gotta be.

[Leo Laporte]

Let me look at what the current engines are that I can choose.

[Jeff Jarvis]

You're telling me you could engage Sexy mode on Perplexity? While you're doing your deep research, I could.

[Leo Laporte]

You know what? All I care about is voices. Now I just want better voices. That's all I really care about. Let me look and see. What?

[Paris Martineau]

Because he's a radio guy, he likes voices.

[Leo Laporte]

I can. So right now I'm using Sonnet Claude 3.7 Sonnet. The default is their own kind of mishmash. It says optimized for fast search. There's sonar, which is an advanced model trained by Perplexity. Probably not trained by, but tuned by my guess. But anyway. 4. 0. 4.5. Sonnet Gemini 2.0 flash. Yeah. Grok 2. Oh, you know what people have been saying, where's Grok 3? Where's Grok 3? That's what it is. That's all. Where's Grok 3? I bet you that's what it is.

[Paris Martineau]

Back to liking Perplexity. Did you see the ad they did dunking on Google?

[Leo Laporte]

I didn't.

[Paris Martineau]

Can you play it? I don't know.

[Leo Laporte]

I can't imagine why they wouldn't want us to. You'd think.

[Jeff Jarvis]

Ads on this show.

[Leo Laporte]

So it's an ad with a Korean.

[Paris Martineau]

Star of Squid Game.

[Leo Laporte]

Oh, he's the Squid Game guy. Oh, now he's in C Severance. Or is it squid? Oh, yeah.

[Jeff Jarvis]

Seems Squid Game esque.

[Anthony Aguirre]

Welcome to your challenge.

[Leo Laporte]

It is Squid Game. They probably spent a lot of money on it.

[Anthony Aguirre]

Answer three questions.

[Leo Laporte]

How do you remove coffee stains from a white shirt?

[Paris Martineau]

It's getting very cold.

[Leo Laporte]

So they're going to freeze him unless he can answer them that. So he launches Pugle Pugle. And Pugle says, oh, no, too many text you. Oh, he's got complexity knowledge. And rinse the stain for excess liquid.

[Anthony Aguirre]

Then detergent and vinegar will get it right out.

[Paris Martineau]

He holds it up to the screen.

[Leo Laporte]

Oh, saved. Correct. How do I make cheese stick to a pizza? Google, of course, said Elmer's Glue. Who was the first Korean actor to win an Emmy award. Oh, he knows the answer to that.

[Paris Martineau]

That's really good, isn't it?

[Leo Laporte]

That's a great ad. I don't know where they would they. Where would they show that? Well, they just showed it on the intelligent show.

[Jeff Jarvis]

That's true. Yeah.

[Paris Martineau]

More free things you do for people.

[Leo Laporte]

This is gonna kill me. I mean, I give them an ad all the time because I use it all the time. I just. I don't know. Thing is, there's no stickiness to any of these. Right. You just move to whatever works best for the moment. And that seems to be a moving target. All the time. Google has announced a new healthcare AI updates for its search. So it's knowledge panels. You know, those knowledge panels now cover thousands more health topics.

[Paris Martineau]

Well, this is. You scroll on down or up, whatever you call it.

[Leo Laporte]

The page goes up, my finger goes down.

[Paris Martineau]

So the company. Oh, no, no, go back, go back, go back. Beginning the story. Oh, jeez, you're so bad. Okay, stop, stop, stop. Twitchy finger. The company unveiled a new feature called called what People suggest which uses AI to pull together.

[Leo Laporte]

Oh no, that's bad.

[Paris Martineau]

I know. Online commentary.

[Leo Laporte]

That's bad.

[Paris Martineau]

With similar diagnosis. Well, I rubbed grapefruit on my tumor and it went away.

[Leo Laporte]

That's what we have Reddit for. We don't need Google to do that.

[Paris Martineau]

That's what I was thinking. Yeah, but if AI is doing it, I don't know, I'm just nervous.

[Jeff Jarvis]

Well, Reddit now has a feature in the app where you can search all the Reddit comments like it's an AI engine, basically. Yeah. Where you can type in app like. And I assume it's also in the browser. I just don't know where you could message it. Being like, I have a grapefruit sized lump on my head. What do I. I guess I don't know what might come up with that.

[Paris Martineau]

Who powers?

[Jeff Jarvis]

What should I do?

[Paris Martineau]

Is it open AI Reddit deal? No, no, it's their own thing.

[Jeff Jarvis]

All the content from Reddit.

[Paris Martineau]

No, but what's the model? What model?

[Jeff Jarvis]

I don't know.

[Leo Laporte]

I think it's theirs, but I don't know. Who knows?

[Paris Martineau]

They made it.

[Leo Laporte]

Seek immediate medical attention. Here are some insights from Redditors on similar experiences and advice. Is exactly why Google's doing this, isn't it?

[Paris Martineau]

Yeah.

[Leo Laporte]

Contact your doctor. Emergency room. It could be a lipoma, it could be a cyst, it could be a tumor. How they would diagnose it.

[Paris Martineau]

That is usually soft and movable.

[Leo Laporte]

This is actually really good. I mean, they're not.

[Jeff Jarvis]

You heard it here first. Leo says go to Reddit for all your medical advice.

[Leo Laporte]

Yeah, yeah, this is great.

[Paris Martineau]

That's how the machine's going to kill us all.

[Leo Laporte]

How is baby made? Oh my God. They're telling me how kids hide your eyes.

[Paris Martineau]

A fascinating and complex process.

[Leo Laporte]

Here's a succinct guide to understanding how babies are made. Well, this is a lot better than.

[Jeff Jarvis]

The Yahoo Answer based on insights from Reddit. That's where I always want my insights.

[Leo Laporte]

On what's the best spatula, because that's really what I use Reddit for, is like recommendations, fish Spatula, Silicone spatula. Wooden spatula.

[Paris Martineau]

Well, okay, look up a fish. But you know what they are giving.

[Leo Laporte]

Look, the wusthof fish turner is. Is the best metal.

[Paris Martineau]

I don't. I've never heard of one before. What's. What makes it a special.

[Leo Laporte]

A fish spatula?

[Paris Martineau]

Yeah.

[Leo Laporte]

You've never seen a fish spatula? No, I was watching my favorite.

[Paris Martineau]

Oh, that's what. That's what you call that. Oh, okay.

[Leo Laporte]

It's funny shape. Yeah, yeah.

[Jeff Jarvis]

This is a fish.

[Leo Laporte]

Oh, you got one. Yeah. There you go.

[Paris Martineau]

Not according to Google. Hold on here.

[Jeff Jarvis]

Right?

[Leo Laporte]

It's a.

[Paris Martineau]

You Google it, you get a very different answer. Google fish spatula. Hard to say.

[Leo Laporte]

By the way, this is the best fish you get.

[Paris Martineau]

The thing with the. With the.

[Leo Laporte]

Yours has tines.

[Jeff Jarvis]

There's also one of those too, but I'm not gonna go get it.

[Leo Laporte]

Okay, so you have a bunch of fish spatulas. This is great. This is. This is all redditors and see, you know. So what's. Oops. What? What? I just typed what an answer. Couldn't quite figure out what the best maple syrup that I. That doesn't come from Canada because they can't get it from Canada anymore.

[Paris Martineau]

Well, you can.

[Leo Laporte]

Yeah, for double the price. Ooh, Escumniac, Late harvest Escominac or Kirkland. Not getting my maple syrup at Costco.

[Paris Martineau]

Costco's.

[Leo Laporte]

Great Chalk Zero Lakanto. No, I'm gonna go with this because it sounds weird.

[Paris Martineau]

Because it sounds like it'd be expensive.

[Leo Laporte]

Yeah, yeah.

[Paris Martineau]

That's what you do.

[Leo Laporte]

Escuminia. Es. Oh, yes. It's only $27 for it, but how big is that bottle? 500 milliliter bottle.

[Paris Martineau]

Okay. Compare that to constant. Go to the Costco.

[Leo Laporte]

This is going to be.

[Paris Martineau]

So that's $27 Canadian.

[Leo Laporte]

Oh, yeah, you're right. Canadian. So I don't know what the. Are there tariffs then? Right. Is there a maple syrup.

[Jeff Jarvis]

There's a tariff on everything.

[Paris Martineau]

Everything.

[Benito]

Can you query just one specific subreddit?

[Leo Laporte]

Yeah, I'm sure you can narrow it down. The funny thing is, Costco's is known for its good quality and value available in large quantities. You need five gallons of syrup. Costco's got it. Here it is.

[Paris Martineau]

20 bucks.

[Leo Laporte]

Actually, look at this. This is the Kirkland that comes in a whiskey bottle. Pure organic maple syrup now at 740 milliliters for 23 bucks. It's Canada grade A syrup de la Belle.

[Paris Martineau]

All right, ask.

[Jeff Jarvis]

I wonder if the Kirkland guys have done a thing on maple syrup. Do you guys know if these people.

[Leo Laporte]

No, can it.

[Jeff Jarvis]

The TikTok group. It's a TikTok group that I think is a band maybe named Morning Trips, I can't recall, but they've become popular on TikTok for this series they do called Kenneth Kirkland, where they try. They compare brand name liquors at first against Kirkland brand liquors in kind of a blind taste test. But they've since branched out to other things and I assume they'd have thoughts on maple syrup.

[Leo Laporte]

I did ask who the best skeeball player is in the world. And while it can't help me with that, it did say there are skeeball leagues. There's a. Oh, there you go, San Francisco. There's high scores. It's just skeeball strategy. You might want to visit R Skeeball for all your skeeball needs.

[Jeff Jarvis]

A lot of it is photos of skeeball.

[Leo Laporte]

Oh, you already know about this.

[Jeff Jarvis]

I mean, maybe I should get on there.

[Leo Laporte]

Just got a skeeball machine. Paint damage. Designed and printed A skeeball game. That's hysterical. Is this the kind of skeeball here that you play?

[Jeff Jarvis]

Yeah, that's the kind of skee ball. Yeah, that's the classic one. Sometimes there's a little thing in front of the holes, but when you get.

[Paris Martineau]

Your stock at the information and you're very rich from. From it and you. Well, let's imagine you buy your mansion on the Hamptons and in the basement. Are you going to buy a skeeball machine of your very own?

[Jeff Jarvis]

Oh, absolutely.

[Paris Martineau]

Yeah.

[Jeff Jarvis]

In my fantasy mansion that I get from the acts of journalism.

[Leo Laporte]

Who was it? That was it lbj that put a bowling alley in?

[Paris Martineau]

That was in the White House.

[Leo Laporte]

It was Nixon.

[Paris Martineau]

That's what went over the pool, I think, right?

[Leo Laporte]

Yeah, he put it over the.

[Paris Martineau]

That became. That became the press room. That's what the press room is now.

[Leo Laporte]

That's the. Where the press room is.

[Paris Martineau]

Yeah.

[Jeff Jarvis]

Whoa.

[Leo Laporte]

What do they do before then?

[Paris Martineau]

I don't know.

[Leo Laporte]

They stand out in the Rose Garden, maybe.

[Paris Martineau]

Well, ask Reddit. I don't know.

[Leo Laporte]

Google plans to release new open AI models for drug discovery. This is good. I'm always looking for new drugs.

[Paris Martineau]

Yep. Yeah. Especially now.

[Leo Laporte]

During a health focused event in New York on Tuesday, Google announced that it's developing a collection of OpenAI models. That's not capital O open.

[Paris Martineau]

That's lowercase open in air quotes.

[Leo Laporte]

Air quotes. Open air.

[Paris Martineau]

As Jason and I like to say. It's open. Ish.

[Leo Laporte]

Open Ish. Open weight. I've Been using open weight lately.

[Paris Martineau]

Okay.

[Leo Laporte]

For drug discovery called TX Gemma. It will be released through its AI developer foundation Pro Health AI Developer Foundation's program later this month. Month can understand regular text the structures of different therapeutic entities, chemicals, molecules and proteins. I mean, I think this is a very promising area. Right. We. We know that it can come up with new proteins and protein folders.

[Paris Martineau]

Oh, it's very promising.

[Leo Laporte]

Yeah.

[Paris Martineau]

Yeah. And you know that there's going to be a huge process. If you discover a molecule that looks promising and you get as far as you can with it, then there's a process that we. Well, I don't know. There was a process that we went through to approve drugs. God knows now. Yeah.

[Benito]

I mean, this is.

[Leo Laporte]

Hey, RFK Jr. You like? This one is. This one's okay.

[Benito]

This is the stuff. This is really the stuff. That's the stuff, right? That AI should be doing.

[Leo Laporte]

Yes.

[Paris Martineau]

Oh, yeah.

[Leo Laporte]

Everybody agrees.

[Paris Martineau]

Yeah.

[Leo Laporte]

Although I did see that they are. They removed mentions of MRNA vaccines from the nih, which is a little.

[Paris Martineau]

Which is also very, very important for cancer research.

[Leo Laporte]

Very. They now have designer. They have designer cancer drugs. They. They do a genetic sample of your cancer and then design a drug that's.

[Paris Martineau]

Aimed specifically at it as a two time loser. Winner. I should say.

[Leo Laporte]

You are a winner. You're still here. You're with us.

[Paris Martineau]

Still here. But yeah. You really, really want advances in these areas.

[Leo Laporte]

Absolutely. Yeah. Yeah.

[Jeff Jarvis]

I'm sorry, guys. You said skeeball and the best skeeball player in the world. I went down a whole rabbit hole.

[Leo Laporte]

Is there one?

[Jeff Jarvis]

I mean, there's contention between Joey the Cat, who's apparent. I've been a very good skeeball player and another person who. Rafi Footerman perhaps, who is upset.

[Leo Laporte]

Queens.

[Paris Martineau]

One from Brooklyn.

[Jeff Jarvis]

We want to see San Francisco. New York City or New York video.

[Paris Martineau]

I want videos.

[Jeff Jarvis]

I'm trying to find it. I've realized that. That there's been a recent upset happened at a local skeeball bar here in Brooklyn, which I didn't realize they were hosting national championships.

[Leo Laporte]

Whoa, you gotta be there.

[Jeff Jarvis]

I do.

[Anthony Aguirre]

I do have to be there.

[Leo Laporte]

One in four big.

[Paris Martineau]

You.

[Jeff Jarvis]

I mean, I wouldn't. There's. I think it's a coed typically. But the ski ball. The skee ball bar that this took place at is Full Circle bar, which makes sense because every time I go there, they have their own ski ball league that we don't participate in because it is. They're good. Insanely good. And it makes sense that this is where the world Champions play. Because I am always like, what? How are you that good? Could, but that's how.

[Paris Martineau]

Video. Is there a video of this guy?

[Jeff Jarvis]

I want to see video Joey the Cat. Search the skeeball kid versus Joey the Cat. Because those are.

[Benito]

Did they bring their own balls or something? Like what's. Do they have probably tools and stuff?

[Paris Martineau]

Oh, that's funny.

[Jeff Jarvis]

That's actually no, maybe it's, it's gotta be. It's got to be standardized. Right here, I'll put this.

[Leo Laporte]

Wait a minute. This is who will win the national skeeball championship.

[Anthony Aguirre]

And the crowd's channel cheering.

[Jeff Jarvis]

That's, that's.

[Leo Laporte]

Is that Joey the Cat also?

[Jeff Jarvis]

I don't know.

[Leo Laporte]

The big mustache. There's the kid. That's got to be Joey. First time I ever played skee ball was full circle.

[Jeff Jarvis]

Full circle.

[Benito]

Okay.

[Paris Martineau]

I want to see the full walk.

[Anthony Aguirre]

In Point Pleasant in New Jersey.

[Paris Martineau]

This is frustrating.

[Leo Laporte]

I mean, look, if you could televise bowling, you could televise this because I.

[Anthony Aguirre]

Liked it and I could play a bunch for five bucks.

[Paris Martineau]

So did I tell you about, about the swifts in typesetting?

[Leo Laporte]

No. What's that?

[Paris Martineau]

Typesetting, one letter at a time, became a spectator sport. It was the. Honest to God, it was the bowling or golf of the day. People, they would have week long tournaments and people would watch, type setters, set type and win money. Yeah, they toured. It stopped because the International Demographical Union forced them to stop because the, the typesetting machine, the latter type that I'm worried about came along and they didn't want a John Henry Man V machine defeat. Whoa, isn't that great?

[Jeff Jarvis]

That's great.

[Leo Laporte]

EFF is concerned about California AB412. They say a bill that could crush startups and cement a big tech AI monopoly.

[Paris Martineau]

He's going to bring us back to the prior conversation.

[Leo Laporte]

The bill demands that creators of any AI model, even a two person company or a hobbyist tinkering with a small software bill build identify copyrighted materials used in training. You'd have to have a list.

[Paris Martineau]

So EFS says this is going to kill the little guys and stop competition with the big guys. Tim, Nick Gebrew on LinkedIn said that's all bogus. If you take ingredients for your stew, then don't you pay for them. Which is going to Paris's argument here.

[Leo Laporte]

Yeah, eff, which you know, I generally trust on these kinds of things. This is Joe Mullen writing, says even for major tech companies, meeting these obligations would be a daunting task for a small Startup throwing on such an impossible requirement could be a death sentence. They'll be forced to devote scarce resources to an unworkable compliance regime instead of focusing on development and innovation and AIs. I mean the EFF's position is that AI training is like reading and it's very likely fair use. So they agree with Kathy Gellis and Jeff Jarvis and I guess I include myself. Although I'm on the fence, I am sympathetic with creator's point of view.

[Jeff Jarvis]

So I'm not sure that it's an impossible standard to have AI developers track and disclose when they use a copyrighted work in their training. Yeah, I think that that's a fairly reasonable thing.

[Leo Laporte]

The problem they say is part of the problem is that searching for a copyright at the copyright office is a very manual thing. It's not machine readable, it's not accessible, it's more like a card catalog than a database. So it's very difficult to identify the provenance of the data that you're reading in.

[Paris Martineau]

And it's going to.

[Jeff Jarvis]

I mean then why have copyright at all?

[Paris Martineau]

Well that's a legitimate question in this day and age which I've written about. Yeah, that's why I suggest something new called credit.

[Leo Laporte]

Right, Tell us about it. Mr. Jeff Jarvis, haven't I told you about this?

[Paris Martineau]

That.

[Leo Laporte]

No.

[Paris Martineau]

My argument is that in a string of, of creativity there are many players who should be noted and potentially rewarded and whose good behaviors toward creativity should be recognized. So if I tell you a story, it inspires Paris to write a poem about it and then Bonito makes a song for that, then Leo puts it in his award winning film. Shouldn't each in that sequence get at least credit? Now whether they get, do they get compensation is a different question. But then you now have the system to enable you to recognize who gave you what in that sequence. It recognizes the creativity is in fact a collaborative effort.

[Leo Laporte]

Corey Doctorow's complaint is that copyright these days really benefits the publishers, the companies, not the individuals. It creates the content.

[Paris Martineau]

Well that's, that's been the case from the beginning. Copyright was not demanded by creators, it was demanded by the industry, by the stationers and publishers and booksellers of England to protect. Licensing lapsed because there was too much publishing in their view and they wanted to control it and they wanted to. And so what's interesting to me is that so, so, so then the publishers said well this is like a God given right that the author has permanent rights to their stuff. So when we buy it we get permanent rights and it went to the House of Lords, where that was turned down. And so what really then copyright is. It's not a granting of rights, it's a limitation of rights. Because the doctrine was that there's a natural right to your creation in public that could last forever, but when you sell it, when you set up those circumstances, you're limited at the time to 14 years. So just interesting little copyright stuff.

[Leo Laporte]

Scooter X is pointing out that for the first time since its launch, the Gemini app no longer requires you to sign in with your Google account to use it. If you go to gemini.google.com and incognito mode or without being logged into Google, you will be able to actually go to the chat interface and query it. So it's free. Free, free. You don't have to have a Google account to do it. I think that's a big move from Google. Good for them. You know what? I think they're worried they're falling behind, that. I don't think they're technically falling behind, but I think in the perception of people, anthropic and OpenAI seem to be completely dominant. And Google is not. Not often considered. Yes or no. Is that just me?

[Jeff Jarvis]

I mean, I think it's not as universally known. Chat GPT is obviously more of a shorthand than Gemini, but I think that's also because Google changes the name of its products, like every other week.

[Leo Laporte]

Yeah, it's part of it too. By the way, if you wanted a sponsor for your skeeball league, I would be glad to kick in. As long as we get a. A logo on your sleeve.

[Jeff Jarvis]

We need to. We need to actually make the jerseys happen this year.

[Leo Laporte]

Yeah, we'll sponsor you.

[Paris Martineau]

How many teams are there in the league?

[Jeff Jarvis]

It depends. Every year at Volo Sport, it kind of changes and it depends which league, like last this year. Last winter, we weren't able to play because we. I think as I mentioned, normally there was like a whole schedule snafu and we ended up getting booted out of our normal day in a. No, I know. And we almost were like, don't you know who we are? We've been doing this for like, we didn't but like the Bushes been doing this.

[Paris Martineau]

If you were men, you do that.

[Leo Laporte]

Do you know who we are with the Bourgeois ski?

[Jeff Jarvis]

It's true.

[Leo Laporte]

Well, if there's anything I can do as a corporate sponsor to support you.

[Jeff Jarvis]

I'll let the team know.

[Leo Laporte]

Il Fogio, an Italian newspaper, says, has published the world's first All AI edition. AI used for Everything, the writing, the headlines, the quotes, even the irony.

[Jeff Jarvis]

Basically.

[Leo Laporte]

Yeah. I can't tell. It's a conservative, liberal daily. I don't even know what that means.

[Paris Martineau]

What's that mean?

[Leo Laporte]

Italy is so interesting. It's part of a month long journalistic experiment. This is a story from the Guardian to show the impact of AI technology on, quote, our way of working in our days. It's only four pages. It was wrapped into the newspaper's own regular edition. I don't read Italian, so I can't tell you how well written it is.

[Paris Martineau]

However, you do read English.

[Leo Laporte]

Oh, is there a translation?

[Paris Martineau]

No. The Independent is launching an AI supported news service.

[Leo Laporte]

Oh, interesting.

[Paris Martineau]

I'm not exactly. I can't figure it out exactly what it is.

[Leo Laporte]

I'll be honest.

[Jeff Jarvis]

AI supported how?

[Paris Martineau]

I'm not sure.

[Leo Laporte]

I often. And you see this. I've mentioned it before. Might see the link in some of our news stories through Perplexity, because Perplexity has a news page which I find quite excellent and I haven't found any errors in it. So I don't know. Interesting stuff. Oh, Google did announce the 9A. We thought they might do that today.

[Paris Martineau]

See, there you go, right there.

[Leo Laporte]

Look at that.

[Paris Martineau]

But also look. And they do link to the stories, but the stories are all basically the same.

[Leo Laporte]

Yeah. Isn't that the funny thing? Isn't that the funny thing? Yeah, because everybody's copying everybody else. We've always liked these A versions of the Pixel phones, so this looks like a pretty nice one. According to AI, this is the most robust A series phone to date. Oh, no. Google said that IP68. Water and dust resistance. Yeah, fine. It's a sleeker design that eschews the iconic camera bar that you find on a lot of the Pixel phones. That big old bar there, I think that's cool. I think there are people who don't want that bar. And some might not worry so much about the camera if they don't have to worry about the bar.

[Benito]

Wait, it's a phone that can lie flat.

[Leo Laporte]

It can lie flat.

[Benito]

It's like there's no phones. Phones can't do that anymore.

[Leo Laporte]

Who'd have thunk it? 499 for the 128 gigabyte model. I guess I could just go to the Google store and get some pictures of it for you. Yeah. You know, I like these Google phones. They call it. This tagline is magic Done Bright because it comes in colors, kids. Purple, pink. We've decided. Lisa and I decided to do everything in hot pink from now on, by the way. Just.

[Jeff Jarvis]

That's great.

[Leo Laporte]

Yeah. It's the new color.

[Benito]

Still a small bump. Still a small camera.

[Leo Laporte]

Bump is a little bit of a wee thing. A little bit. 30 plus hour battery life. They claim up to 256 gigs of storage. So there's not.

[Paris Martineau]

Is there a price?

[Leo Laporte]

Yeah. 499 for 128 gigs.

[Paris Martineau]

Wow. That's pretty good.

[Leo Laporte]

Yeah. These are good phones. We've always thought that the A series phones. So Google releases phones in the fall just around the same time Apple releases its iPhones. Sometimes October, sometimes August. And then around springtime releases the A version, the less expensive version of it. Gemini built in AI Creative Canvas. Look at that tiny chihuahua enjoying a bully stick at home. Oh, hey. I think I might be wrong, but I think that sloth is enjoying some cacio e pepe. Oh, yum yum.

[Jeff Jarvis]

It looks like ramen.

[Leo Laporte]

Actually. It does. Never mind.

[Paris Martineau]

So when I had my. When I had my colonoscopy last week. I know you're not going to want to know any more about that. My wife tried to be good for me and. And went and bought things I could have on that day where you're supposed to have nothing but clear things. And she got pho broth.

[Leo Laporte]

Yeah.

[Paris Martineau]

I think it's. Isn't it pronounced?

[Leo Laporte]

It's pronounced pho. Yeah, but it's spelled, so.

[Paris Martineau]

I'm no FU fan.

[Leo Laporte]

I learned you're not a fu fan.

[Paris Martineau]

I'm not a foot fan.

[Leo Laporte]

I like pho. You don't like.

[Paris Martineau]

I see why one might know how.

[Benito]

To eat it, though. There's. There's a technique.

[Leo Laporte]

Well, there's no noodles in his foot. Yeah.

[Jeff Jarvis]

But I'll pause. It's not for you. Thank you. Thank you, thank you.

[Leo Laporte]

How about this? From Cornell University. An AI ring that tracks spelled words. Not. It's the spelling, not the asl. But it does do the spelling.

[Paris Martineau]

Well, finger spelling is asl. Right.

[Leo Laporte]

Well, there's finger spelling and then asl.

[Paris Martineau]

Is that part of asl?

[Leo Laporte]

Yeah.

[Paris Martineau]

Okay.

[Jeff Jarvis]

How does it figure out what the other fingers are doing? And also isn't face because it's smart.

[Paris Martineau]

Enough to take over mankind.

[Jeff Jarvis]

Isn't face like. Isn't your facial expressions a big part?

[Paris Martineau]

No, it's just the spelling letters. Just the letters.

[Leo Laporte]

That's why probably ASL is much more gesture.

[Paris Martineau]

Yeah. Much richer.

[Leo Laporte]

Yeah. But it does recognize continuous finger spelling in American Sign Language just by wearing a ring. A ring with a rather large circuit board attached, but. Okay. I think that's cool. There's Some really interesting.

[Jeff Jarvis]

Leo, I do think that what you should do is you should get all of these AI products, wear them and then take a photo like that with the ring, the bee, the glasses, the pin, everything.

[Leo Laporte]

Here's the question, though. These days, everything's like, this is the OURA ring. It's probably using some AI, as is the Apple watch, to understand the movements and stuff. Like, the watch knows if I start rowing, it says it looks like you're rowing. So, you know, I mean, there's AI in all of this stuff now. That's the point.

[Benito]

But you wouldn't have called that AI before transformer technology.

[Leo Laporte]

No.

[Paris Martineau]

Yeah.

[Benito]

So why are you calling it AI now?

[Leo Laporte]

Well, the way they did it is they brought in a lot of athletes doing a lot of different movements and they tracked those movements. Movements and then created a model based on those aggregate movements that they don't know what's going on, but they notice that my hand's going like this. And they say usually that correlates. I think it's probably much like an LLM. It correlates to a certain kind of exercise, I guess.

[Jeff Jarvis]

Kevin Roose, Marketing.

[Leo Laporte]

That's what marketing. That's the difference. Marketing. Kevin Roos agrees with our guest earlier today. Powerful AI is coming. We're not ready. We know that he is an accelerationist. He loves his AI. You know what, His AI loves him.

[Paris Martineau]

That's. That's the other thing. Yes.

[Leo Laporte]

Yeah. He says, I believe that when AGI is announced, there will be debates over definitions and arguments about whether or not it counts as real AGI. But these won't matter because the broader point that we are losing our monopoly on human level intelligence and transitioning to a world with very powerful AI systems in it will be true. You don't. You really believe that that will never happen, Jeff?

[Paris Martineau]

No, it won't. I'll say. Okay, long, but I think, but again, it's the wrong question because it can already do. Do a huge list of things way better than we can ever do it.

[Leo Laporte]

Absolutely right.

[Paris Martineau]

That's what we pay attention to. And this idea, it's the, it's the, it's the hubris of our species.

[Leo Laporte]

Well, that's why I did like Anthony's definition because he. The A for him in AGI's is autonomous. And I think that is a very important. That's something you can measure like an autonomous.

[Paris Martineau]

What's also necessary for his nightmare.

[Leo Laporte]

It is necessary for his nightmare.

[Paris Martineau]

Controllable because it's uncontrollable.

[Leo Laporte]

That's right. So I think that that is something maybe to add to the definition. I don't know.

[Benito]

I think there's even a bigger question there of, like, is general intelligence even a real thing?

[Leo Laporte]

Well, it is a meaningless thing because as we.

[Benito]

Artificial intelligence, as Jeff just pointed out.

[Leo Laporte]

Some people know more than other people. People. There is no. Right. So. And by the way, Rus quotes Dario Amode, who is the chief executive Anthropic, as he doesn't like the term AGI, but agrees with the general principle. He says last month, we're a year or two away from having a very. This is more something to his liking. A very large number of AI systems that are much smarter than humans at almost everything.

[Paris Martineau]

What's smarter? Right.

[Jeff Jarvis]

I mean, it's what humans. Are you measuring again?

[Leo Laporte]

Well, we can't. Yeah, I mean.

[Paris Martineau]

Yeah, that too. There's.

[Leo Laporte]

There's still a debate over IQ tests. I mean, are those meaningful? They measure anything?

[Anthony Aguirre]

No.

[Jeff Jarvis]

I mean, they're probably not. I feel, like, very debunked.

[Leo Laporte]

Right.

[Paris Martineau]

Yeah.

[Leo Laporte]

So if you can't measure human intelligence, how are you going to measure machine intelligence? Right. I think we'll know it when we see it. It. That's why even though the Turing Test is somewhat debunked, there is some value in it. The point of the original Turing Test, proposed by computer scientist Alan Turing, was many years ago. Well before AI was, if a human can't distinguish in a conversation, a blind conversation between a human and a machine, which is the human, which is the machine. That's the Turing Test. Last week, we talked with Ray Kurzweil, who had a kind of a more advanced Turing Test that he felt would be more reflective. But he's always said, and I think it's true, if you can't tell the difference, what does it matter what's going on internally?

[Jeff Jarvis]

Because then who do you blame when someone goes wrong?

[Leo Laporte]

Well, who do you blame? If you have an accident, you blame.

[Jeff Jarvis]

Okay, that's not what I'm talking about. I'm just saying if someone. If the person I am talking to.

[Paris Martineau]

Liability.

[Jeff Jarvis]

The question I call a doctor and they give me bad medical advice, place that harms me. If I'm looking for the person who has done me wrong, the answer is that Doctor. If the doctor was a large language model pretending to be a doctor, then who do I blame for my grapefruit?

[Benito]

Right? It's the liability. Who's liable?

[Leo Laporte]

Well, Doctors make mistakes. AIs make mistakes.

[Benito]

Yeah, but then the doctors don't think.

[Leo Laporte]

That makes it better, that you can. You can sue the doctor, but you can't sue the AI.

[Benito]

There's recourse is the thing. There's recourse. There's no recourse with an AI.

[Paris Martineau]

Oh.

[Leo Laporte]

Where did you use this? Did you use it at your doctor's office? Did you use it? I mean there's recourse. There's somebody who was running the machine that gave you the bad diagnosis. I just, I mean, I think it could do the same thing. A human will do the same thing. Does it make it better that you can sue them? I don't know. You still have a lump the size of a grapefruit.

[Jeff Jarvis]

I mean, I think that liability is a real and concrete way to think about this. It doesn't mean it's the best way to think about it, but it is one of the hurdles that we're going to have to come up against. And we're going to have to come up against honestly right now, if not very soon. These tools are being used in medicine, in health insurance, in the government. You know, who is to blame when the AI cuts an entire department that was responsible for important national security issues?

[Leo Laporte]

Well, are there countries where you can't sue a doctor for malpractice? I bet there are.

[Jeff Jarvis]

I'm pretty sure that malpractice isn't as big of a thing outside of malpractice. Litigation isn't as big of a thing outside of the U.S. i think it's like a U.S. specific thing because we're a very legal liability. Like most of the answers to who is to blame in the US come back to lawsuits because of the way that this country.

[Paris Martineau]

Well, when you have state run health care, when you have single payer, then you know, in the UK it's the National Health Service and probably you can't sue it as easily as you could here a company.

[Leo Laporte]

Yeah. I mean in the United States you can't sue governments, right?

[Paris Martineau]

Well, yeah, you can't. Sure you can.

[Leo Laporte]

You can. In general, governments are protected against laws.

[Jeff Jarvis]

Okay. Someone in the chat is saying how ridiculous is the statement that we shouldn't use a method of solving a problem if we can't attack them if it does wrong. That's insane. That's not entirely. What I'm saying is, I'm saying when we are handing over large segments of the decision making process to autonomous machines, to intelligent machines, you might say we need to be thinking about the consequences of that and what we would then do if that machine produces outputs we don't like. Because we can't just assume that Everything is going to go super well. That's not, I think a great idea when we're thinking about large scale systematic changes in our society.

[Paris Martineau]

I couldn't agree more. And I think we start with so called autonomous cars and test Tesla. I cannot believe that they are allowed on the road with what they have. Yeah.

[Benito]

Wait, what happens now? When if a Waymo crashes into somebody who, what happens?

[Leo Laporte]

You sue Google.

[Paris Martineau]

You sue Google.

[Benito]

Okay.

[Paris Martineau]

Speaking of which, there's a new Waymo in San Francisco.

[Leo Laporte]

They're testing it right now.

[Paris Martineau]

Yeah, I hitched a ride. Line 123. Yeah, this is the one without any steering wheel.

[Leo Laporte]

This looks like the Zooks. Oh, it is Zoox. This is Amazon's.

[Paris Martineau]

Oh, Amazon, sorry.

[Leo Laporte]

Yeah, yeah, this is the Zoox. This has been around for a while, but I'm, I'm. Yeah. Why put in a steering. Why have a front seat if you don't have a driver? So Zoox is from Amazon and it is now testing in, in San Francisco. Waymo's been around San Francisco for some time. I think the Zooxes are kind of cute. This looks exactly like the, the taxis in Westworld. Of course.

[Paris Martineau]

So. So the three of us get in one for our 24 hour festival that we want to do.

[Leo Laporte]

Okay.

[Paris Martineau]

Do you. Who, who tries. Who's willing to sit facing backwards?

[Leo Laporte]

I don't care. I'll sit back facing backwards. Who's willing to go over a bridge.

[Paris Martineau]

In that?

[Leo Laporte]

You don't like sitting backwards?

[Paris Martineau]

Not. Not.

[Jeff Jarvis]

Yeah, not in a moving vehicle.

[Paris Martineau]

Yeah.

[Leo Laporte]

Really doesn't bother me at all. Really. I often take a train.

[Jeff Jarvis]

I was gonna say, if you're getting on a train, you're like, oh yeah, the backward seat. You're not a little bit like. I'd prefer the front, the forward facing seat.

[Paris Martineau]

I do too.

[Leo Laporte]

Yeah, well, there's others.

[Jeff Jarvis]

Not saying I won't sit in the backward facing. See, I'm just saying if I have a prep. If a choice, I'm choosing forward seat.

[Paris Martineau]

Yeah, well, especially in this case because you kind of want to know where it's going so you can at least scream when you know you're gonna hit something.

[Leo Laporte]

Don't you want to know where you've been?

[Paris Martineau]

What about.

[Jeff Jarvis]

What about respecting the past? What about respecting from where we've come?

[Leo Laporte]

Yeah. By the way, good news for Elon. X is now worth $44 billion. Exactly. What he paid for was 10 billion in September. You know, it really helps when you own a government. You can really not so much.

[Paris Martineau]

Tesla.

[Leo Laporte]

Yeah, Tesla's not doing well.

[Paris Martineau]

He's lost, let's see. Lost 250 billion. I think it is something like that. Yeah.

[Leo Laporte]

And that's a hard one to stop because it's not just the US that that's happening. It cracks me up though that conservatives who have been anti New Green deal and anti EV for so long are now buying Teslas. Trump's hawking them on the front lawn of the White House. We live in a very strange world.

[Paris Martineau]

Let'S just say that ever.

[Leo Laporte]

Right? Right. You can buy the 560 pound Twitter logo right now. It's up for sale for auction.

[Jeff Jarvis]

Guys, we should go in. We should go all in on it. We could each take apart.

[Leo Laporte]

Where do we put it? You could put it 12 behind you.

[Jeff Jarvis]

Will it fit?

[Leo Laporte]

Yeah, it would fit on the wall behind me. Current bid is 27,500. So you'd have to bid $30,250 to beat the current bid. But it's going to go on for a while, another 24 hours, another day.

[Jeff Jarvis]

Remember when daos were buying stuff like that and they would give one piece to everybody?

[Leo Laporte]

Yeah. Maybe there is a dao going to buy it. The Bluebird logo that was on the front of the Market Square headquarters.

[Paris Martineau]

Market Street.

[Leo Laporte]

Yeah. I don't know why it says Market Square. The auctioneer has got it wrong. Yeah, the headquarters were square though. Maybe that's what he means. Market Square headquarters. It's a they. Oh, you know what? Somebody bought this in September or of 2023. Elon had an auction. The rebranding auction. Somebody bought it, now they're selling it off.

[Paris Martineau]

Oh, I want.

[Leo Laporte]

It's an investment.

[Jeff Jarvis]

Oh, they're flipping it. They're flipping the bird. They're flipping the bird.

[Leo Laporte]

Okay, we have a show title. Thank you for your submissions. Mark Andreessen wants to shut down all higher education.

[Paris Martineau]

Yeah, this was freaky.

[Leo Laporte]

What the hell's wrong with this guy?

[Paris Martineau]

This guy is really over the ass.

[Leo Laporte]

This guy has just gone off the deep end.

[Paris Martineau]

And in exchange with Lex Fridman, who just drives me bananas.

[Jeff Jarvis]

Does he have kids? See if college age kids.

[Paris Martineau]

Andreessen. I don't know.

[Leo Laporte]

Andreessen said there's no way to fix higher education in America without replacement. And there's no way to replace them without letting them fail. This is the most obvious conclusion of all time. How come you don't understand this? What happens in the business world when the company does a bad job? It fails another company takes its place. That's how you get progress. So can we explain to Mark that a university Is not a business.

[Jeff Jarvis]

What about all of your. The companies you've invested in, Mark, that haven't made money for years? Part of the deal.

[Leo Laporte]

Just because he has a note. He is a notorious nitwit. I'm sorry.

[Paris Martineau]

Yeah, he is 8 year old son. No, now 10. That was two years ago. So 10.

[Leo Laporte]

Well, I wonder if he goes to college.

[Jeff Jarvis]

It's often my. I often find whenever people are making the argument that colleges need to be shut down, they're brainwashing the kids. You look at their kids. Kids and see if they've recently gone to college and whether or not they speak to them.

[Leo Laporte]

Are they brainwashed? Yeah. Well, let's not forget that Marc Andreessen has money because he went to the University of Chicago, Illinois at Urbana, the University of Illinois in Urbana, at the national center for Supercomputing Applications where he wrote the first browser. What was it? Was it Mozilla? Was it Netscape? Netscape Mosaic? That's what it was. You're right. And thanks to all the money that was provided that national center for Supercomputing Applications by the federal government. Again, notorious nitwit. This is unfortunately a problem with people who fall into a lot of money and then think they're somehow geniuses. I do not have that problem. I am proud to say I have no money and I'm an idiot.

[Jeff Jarvis]

Not a nitwit, but a twit.

[Leo Laporte]

There is. Yes, exactly right. There is an auction that I am intrigued by. The Free Software foundation is having a silent memorabilia auction, which, I mean, I don't know what they're auctioning off, but I feel like it could be like hairs from the beard. I don't know what it is. I don't know what it is. This is. It'll be interesting to see what they're selling first.

[Jeff Jarvis]

10 items are up.

[Leo Laporte]

A variety of plush animals that greeted visitors in his former Boston office.

[Jeff Jarvis]

The Dynamic Duo, The GNU and the Penguin.

[Leo Laporte]

Oh, the gnu. The original GNU head. It's autographed by RMS Autograph. See if I don't get a little piece of Richard Stallman. I don't know if there's a Norbert Wiener Award for. They're auctioning off their Norbert Wiener Award. A pair of GCC GNU letter drawings by Gary Taurisi Studio. That's the COVID of the book. Using gcc, the GNU Compiler reference Manual. The Internet hall of Fame medal awarded to Richard Stallman. You shouldn't be able. Oh. A legendary katana sword given to Stallman by Friends of the Fsf Ooh. The sword resembles the one drawn in the XKCD comic and can be used as a weapon in the fight for computer freedom.

[Jeff Jarvis]

That's really good. It's good. They specify that.

[Leo Laporte]

And there he is. Richard Stallman with his katana. Ninjas are attacking him. A night of blood I've long awaited. Free software will carry on for a.

[Paris Martineau]

GNU Don Linus Torvalds. I hear he sleeps with nunchucks.

[Leo Laporte]

That's hysterical. All right, all right, all right. Right. All right. Let's see. We need to take a break. You're watching Intelligent Machines, brand new episode 811 with our guest, Anthony Aguirre from the pro human sector. I don't know, what was it called? The Institute for the Future of Life.

[Paris Martineau]

Institute.

[Leo Laporte]

Future of Life Institute, Right? Something like that. Probably Paris Martineau of the information and Jeff Jarvis, who does not believe in the future of life. So that's his problem. That's all I can say. If you're not yet a member of the club, I would like to invite you to join Club Twit. Without the club, there would be no scintillating conversations like this. The club also hosts a bunch of stuff inside the Discord that are, you know, want. Tonight at 6pm, just a couple hours from now, less 6pm Pacific, it's Micah's crafting corner. We've got a Chris Marquardt photo segment coming up. Stacy's Book Club. We're going to do another coffee segment with the coffee geek, Mark Prince. All of those in the clubhouse, which is our wonderful Discord. Plus you get ad free versions of all the shows. I think if you hear all of that, you might say, well, that must cost 30, 50, $100 a month. But no, it's only $7 a month. And it helps us put on all the programming we do. It makes a big difference to our bottom line. It covers all of the expenses that advertising does not. And that is a significant portion, sad to say. If you're not yet a member, we would love to have you in the club. Twit. TV Club. Twit. What else? I think we talked enough about gdc. Did you see anything else you wanted to.

[Paris Martineau]

No.

[Leo Laporte]

No. Any other stories? Blue sky made more money selling those. Those T shirts. Jay Graber's T shirts mocking Mark Zuckerberg than it has selling custom domains.

[Jeff Jarvis]

They're a T shirt company now.

[Paris Martineau]

Pivot. Pivot. It's like an everyday T shirt.

[Benito]

It's like how every band is actually a T shirt company.

[Paris Martineau]

Yes.

[Jeff Jarvis]

Yeah, yeah.

[Paris Martineau]

Isn't the Google story this week a big deal.

[Leo Laporte]

Oh, yeah, we should talk about that. The federal government is of course still going after, after Google. They didn't drop that. That's what we want to talk about.

[Paris Martineau]

No, I'm talking. They're buying cybersecurity Firm Wiz for $32 billion.

[Leo Laporte]

Yeah, I mean if this were this week in Google, I might do it.

[Paris Martineau]

Well, what's interesting, what would you like.

[Leo Laporte]

To say about it? It's a huge.

[Paris Martineau]

They tried to buy it last year for like 23 billion or something. I can't scroll this right now, so I don't know how much it was. And they didn't because they feared antitrust.

[Leo Laporte]

Ah. But now they're not worried about it.

[Paris Martineau]

Now they're not worried about antitrust. Or should they be? Because supposedly the Justice Department is still going after the FTC or whoever it is.

[Leo Laporte]

No, well, that's because the trial. No, the trial is over. They lost. So the judge gets to, you know, decide what to do. And the ftc, I guess the lawyers there are still suggesting remedies.

[Paris Martineau]

Well, they just fired the two Democratic FTC commissioners.

[Leo Laporte]

But I think once you, once you've lost a trial. Well, I mean, I guess the FTC could say, well, we don't want any remedy, thank you, but the judge is going to get to decide.

[Paris Martineau]

But so now they're trying, now the price went up, but they're trying to buy it because they've got tons of cash. And so they're buying this for their. So I think that's a big deal.

[Leo Laporte]

And then also purchase filthy. Notably financial crimes.

[Jeff Jarvis]

The fee if they back out on the deal now is astronomical. Like the walk away fee is crazy big, if I recall correctly.

[Leo Laporte]

Well, all this is expensive. They're spending a billion dollars to keep the employees. It's an Israeli cyber intelligence company, I guess Security. Cyber security, yeah. I mean it's. These are. This is the largest acquisition in Google history.

[Paris Martineau]

Yeah, that's a big deal.

[Leo Laporte]

When it bought Motorola, it was only 12 and a half billion.

[Paris Martineau]

Right. How'd that go?

[Leo Laporte]

Mandiant was 5 billion. Yeah. Really?

[Paris Martineau]

So Alphabet, meanwhile, spins off Starlink competitor Tara. Did you know about that? I didn't even know that existed.

[Leo Laporte]

No. And I would love to use a Starlink competitor. We use Starlink as our backup Internet access should Comcast go down? Because it does. Or it had for a while when we first started streaming from here. It's actually been pretty solid since knock wood right now. Well, but I got this satellite dish on the roof. But I would love to look at an alternative. Absolutely.

[Paris Martineau]

So this is a terrestrial.

[Leo Laporte]

That's the difference. That's the difference. Internet from terrestrial lasers. And this was an Alphabet X. Right. X incubator, Moonshot. The company currently has two dozen employees, so it's not huge. It operates in 12 countries, connecting the capital of the Democratic Republic of Congo to the Coachella festival. Oh no, wait a minute. That was two separate installations.

[Paris Martineau]

That's a long way to go.

[Leo Laporte]

Although I think that would be a good. Useful. We've realized over time that for a number of things we create, there's a lot of benefit to landing just outside of the Alphabet membrane. This from Astro Teller. X's, captain of moonshots. Is that what it says on his card? Hi, I'm Astro Teller, Captain of moonshots. They're going to be able to get connected quickly to market capital. Oh, that's it. That way, raise money and scale faster. The current tech involves firing a narrow beam of light from one traffic light sized terminal to another transmission of up to 20 gigabits a second over 20 kilometers. So it isn't. It's nothing like satellite. It's. It's. But there are other satellite competitors, by the way. Coming.

[Benito]

Does this thing need line of sight or something like.

[Paris Martineau]

Yeah, it's line of sight, but it's what it's for.

[Leo Laporte]

So you put it on a tower.

[Paris Martineau]

Yeah, that's the thing. So like microwave.

[Benito]

Like a bird sitting on it is gonna mess things up.

[Leo Laporte]

Don't. No birds allowed.

[Paris Martineau]

It'll fry the bird. So my 8th grade honors. You've heard already about my 6th grade science project. Electronic bungos?

[Leo Laporte]

Yes.

[Paris Martineau]

In my 8th grade honors science class we all had a project and I was trying to work on transmitting information with light.

[Leo Laporte]

That's great. Did you succeed?

[Paris Martineau]

No, I failed.

[Leo Laporte]

Oh, were you blinking the light? What were you doing?

[Paris Martineau]

Well, I had a cell and I had a light bulb, but I couldn't figure out what to do about the ambient light.

[Leo Laporte]

And so you needed lasers, that's why.

[Paris Martineau]

Well, yeah, I didn't have.

[Jeff Jarvis]

Could have gone to a very dark room, you know.

[Paris Martineau]

Yeah.

[Leo Laporte]

Amazon is killing off. It was any boy.

[Paris Martineau]

I was a geek early on.

[Leo Laporte]

I know, I'm impressed. I had no idea you were much geekier than I was.

[Paris Martineau]

Yeah, what happens. I even was. I was an AV guy too. I mean, wow. Paris would have made. Paris. Being the cool kid would have made fun of me.

[Jeff Jarvis]

No, I was. I was a nerd and a geek.

[Leo Laporte]

I was a theater kid.

[Jeff Jarvis]

I was gonna say I was a theater Kid. Yeah, I liked anime. I wrote fan fiction. I was not cool.

[Leo Laporte]

Okay, we didn't have anime when I was in high school. That had not yet been invented.

[Paris Martineau]

Wait a second, Leo, we got over this. What did you. What? A few roles, please.

[Leo Laporte]

I was Sheridan Whiteside and the man who Came to Dinner. Oh, so that was a great role.

[Jeff Jarvis]

Did you come to dinner?

[Leo Laporte]

Yes. The whole idea was he was a famous New York critic, famously cantankerous, who came to dinner at a house, slipped on the stairs, broke his leg, and was forced to recuperate at the house where he became a very annoying guest over a period of many weeks. My first line was off stage. In fact, it's the beginning of the play. It's very famous, funny play. My first line was, as I remember, you driveling cow, you have the touch of a sex starved cobra. And then they wheel me in the whole play. I was in a wheelchair. I also played, as I remember, I played the hero and as yous Like It. Dubois, I think was his name. What else? Yeah, no, I did a few. Oh, oh, and I was also Richard Burton in the Night of the Iguana.

[Paris Martineau]

Whoa.

[Leo Laporte]

Which required a considerable amount of drinking, which is a big deal for a 14 year old.

[Paris Martineau]

I was Walter Hollander, the Jewish caterer from Newark, New Jersey in Woody Allen's Don't Drink the Water.

[Leo Laporte]

Oh, that's great. How fun. Did you play it for comedy?

[Paris Martineau]

Yes. But as it turned out, it was the lead. I'd never been in anything. I got it accidentally because I accidentally.

[Jeff Jarvis]

Get a lead role.

[Paris Martineau]

So I filled out the form of what I had moved from. From New Jersey to New York. And I filled out the form where you've done before. And so I mentioned productions. I made very clear to say I was in lighting. I was the geek. I was the. I was the. The boring kid. And the drama coach thought that I was like a star in those productions. And so I did a good audition and I got the lead. I was. I was really crappy.

[Jeff Jarvis]

Well, do you think it's just that there weren't any men that wanted to be in the production? Because that's a problem some schools have.

[Paris Martineau]

So there was a guy who was in it who became a movie star. Namir Al Kadi.

[Jeff Jarvis]

Whoa.

[Paris Martineau]

Was in my class. He was a. He's a movie star. Yeah.

[Leo Laporte]

That's pretty cool.

[Jeff Jarvis]

You could have been a laser inventor. You could have been a movie star. Instead you chose to be here on this electronic percussionist.

[Leo Laporte]

I'm actually, I'm looking at a lot of the man who came to dinner. I had forgotten this. He was a radio personality. An acerbic New York radio personality. So this may be where it all began.

[Paris Martineau]

We have found your Leo.

[Leo Laporte]

Wow, I forgot that I thought he was a theater critic. No, he was an acerbic New York radio personnel.

[Paris Martineau]

All that and the voice. Yeah.

[Leo Laporte]

Great dribbling cow. You have the touch of a sex tone.

[Paris Martineau]

Cobra's rosebud. Yep, we're there.

[Leo Laporte]

You just. We just found my rose.

[Paris Martineau]

You found. We found the unified theory of Leo.

[Leo Laporte]

You're watching this week in intelligent machine machines.

[Paris Martineau]

Whoops.

[Leo Laporte]

Almost said Google said it. I almost said it. I knew I would at some point. We're gonna take a little break. When we come back, it's pick of the week time. Paris and Jeffrey. All right, Paris Martineau, kick us off with your pick of the week from the New York Post.

[Jeff Jarvis]

Nope, I put the wrong. I put the wrong link there. That's something work.

[Leo Laporte]

Hold on. Oh, I'm not secret. She put in a work link.

[Jeff Jarvis]

Listen, sometimes you got a lot of links in the clipboard before.

[Leo Laporte]

Are you talking about the Broadway show oh, Mary? When you say oh.

[Anthony Aguirre]

I am.

[Jeff Jarvis]

Yes, I. So last night I went to go see the Broadway show oh Mary, which just opened with a new person playing Titus.

[Leo Laporte]

Oh, my God. I love.

[Jeff Jarvis]

He was fantastic in it. And I just really recommend. If anybody's coming through New York and wants to see the show. Omiri was really good. Titus is the lead for, I think, the next couple weeks, the very least. And then it's gonna be back to.

[Leo Laporte]

Cole Escola, who wrote it and created the part originally.

[Jeff Jarvis]

Yeah. And I'm gonna go back and.

[Leo Laporte]

Is it hysterical? I've seen him like on the talk shows in his Mary Lincoln gar. But it's hysterical. But so it's a. It's a one person show about Mary Lincoln.

[Jeff Jarvis]

Ish. It's technically they have a cast of five, but Cole, there's a couple of other supporting characters, but they don't really play big of a role because it's all about Mary Todd Lincoln. And it is, I would say, very loosely related to history or Mary Todd Lincoln at all. It is basically using Cole, by the way.

[Leo Laporte]

Way. The original. Yeah.

[Jeff Jarvis]

It is basically a gay romp that is a mishmash of queer and gay culture in Mary Todd Lincoln. Like part of one of the core tensions is that Abraham Lincoln is secretly gay and trying to fight off his repressed homosexuality. Mary Todd Lincoln is a manic depressive alcoholic who can't stop lying about everything. Kind of what if Mary Todd Lincoln was a real housewife is kind of a good way to think of the show. That's hilarious.

[Leo Laporte]

What a premise.

[Paris Martineau]

My mother insisted that she was Mary Todd Lincoln reincarnated.

[Leo Laporte]

Oh, wow. Not somebody to emulate. She was kind of a horrible person.

[Paris Martineau]

Yeah.

[Leo Laporte]

I wanted to see this so badly. And I love Titus Burgess. He was in the Unsinkable.

[Jeff Jarvis]

Unbreakable Kimmy Schmidt.

[Leo Laporte]

Unbreakable Kimmy Schmidt.

[Jeff Jarvis]

It's really quite. I mean, you get to see all of his skills, really.

[Leo Laporte]

He was one of his funniest. I follow him on Broadway before, but.

[Jeff Jarvis]

More in, you know, supporting or cast, like, roles. And so it's really great. Was great to see him in his first, like, leading Broadway role. He killed it. It is a incredibly life, physical role. It's very funny. It obviously requires a lot of acting, but it also requires some singing and cabaret. It's not a musical, but that comes into play in some parts of it kind of unexpectedly. And he really shone.

[Leo Laporte]

So would recommend and it would drive the administration crazy because it's a drag show.

[Jeff Jarvis]

That's true. And I think I heard that. I don't know what, this could be fake news, but I think I heard that part of the reason why Cole, the original author and person who played Mary, why he. Why they and the whole Broadway cast are all coming back in April is because I think they're going to be filming it for streaming. You might be able to see it from the comfort of your home.

[Paris Martineau]

All.

[Leo Laporte]

I would love to. I would love to. How funny is that? I'm so jealous. See, this is why I wished I lived in New York City, because I would go to the theater all the time.

[Jeff Jarvis]

It's fantastic.

[Paris Martineau]

It's a theater, kid.

[Leo Laporte]

Yeah. There's something about a live performance that is just. There's nothing like it. It's just amazing.

[Jeff Jarvis]

It really is fantastic. It was also just. I mean, this is the first time I think I've seen someone on Broadway. It was literally the opening night for Titus on Tuesday. I don't know. I mean, I think I had just. I had just. Just seen it come across my Twitter feed that Titus was going to be playing this. So I opened it up and was like, oh, tickets are kind of cheap. So I nabbed them, but then they spiked, so I.

[Leo Laporte]

Now you can't get them? Yeah, yeah.

[Jeff Jarvis]

The tickets are really expensive now. But it was incredibly. I felt like moving just to see at the end the curtain call of, like, someone experiencing a, like, five minute standing ovation on their first night. In Broadway. Really brought tears to my eyes.

[Leo Laporte]

Yeah.

[Jeff Jarvis]

I don't know. Go visit your local theaters, guys.

[Leo Laporte]

A great institution. Theater is incredible. It's amazing. It's amazing. I guess that means it's your turn, Jeff Jarvis.

[Paris Martineau]

So, you know, the other day, I was wondering, whatever happened to Mike Arrington?

[Leo Laporte]

Oh, my God. Whatever happened to Michael Crunch? I'm not exactly sure he's doing all right.

[Paris Martineau]

Yep. He bought a house in Miami beach for $60 million.

[Leo Laporte]

What? Oh, crypto, baby.

[Jeff Jarvis]

Oh.

[Leo Laporte]

Oh.

[Paris Martineau]

TechCrunch founder turned crypto investor.

[Jeff Jarvis]

Boy.

[Leo Laporte]

The house was built by a real estate mogul for seven and a half million dollars in 2020. He spent four years building it. They moved into the home last fall, Wall. And sold it without a realtor to Mike Arrington for 60.

[Paris Martineau]

Well, they saved on the commission.

[Leo Laporte]

Yeah. The property was not on the market at the time of sale. Oh, this is so Mike Arrington.

[Paris Martineau]

So Mike.

[Leo Laporte]

Mike just said, I'm buying your house. He got a call from a real estate agent with another overture. Apparently, he was getting one every week. He was about to turn him away when Karp suggested he throw out a number. $65 million, said the realtor, which would net him a significant profit, according to the Wall Street Journal. And Mike said, 60. Wow. But they just.

[Paris Martineau]

Five bedrooms, two staff bedrooms.

[Leo Laporte]

You gotta have staff bedrooms.

[Paris Martineau]

You have to. Fourth of an acre. Pool, elevator, gym with sauna and steam room. Picture Mike in the. On it.

[Leo Laporte]

I made a big mistake throwing him off of our shows.

[Paris Martineau]

Yeah.

[Anthony Aguirre]

What? What?

[Paris Martineau]

What did he do?

[Leo Laporte]

He accused me of favoritism because I got a review phone of the. It was the Palm Pre, and I really liked it. And he said, did they. Did they give that to you? And I said, yeah, for review. I'm returning in two weeks. Said, huh, huh. So finally, I just swore at him and threw him off. Big mistake. I couldn't find a place to stay in Miami if I had just stayed friends. I saw him a couple of years later at Loeweb in Paris, and I said, let me shake your hand. He was holding court. He was sitting high atop a pile of cushions, so he was about 8ft in the air. It was really weird. Weird. It was a fee in the air. And I reached up and he said, no, no, I'm not shaking hands these days. Last time I saw.

[Jeff Jarvis]

Not shaking hands these days.

[Leo Laporte]

I bet he has a pillow throne in his new home.

[Jeff Jarvis]

However, kind of a whole pillow room.

[Paris Martineau]

It's all. It's a wonderful accident story because I talked to him early on. I haven't talked to him in years. I talked to him early on in his business in TechCrunch. And TechCrunch was. It's like Craigslist, but. But rich in that he was just making lists for the sake of his own work. And TechCrunch became a thing that he didn't anticipate.

[Jeff Jarvis]

Huh. Interesting.

[Leo Laporte]

He did a good job. I mean, tech. I mean. But I don't think he sold it for that much, so. And he became an investor. I think this is really. He fortuitously invested in Bitcoin.

[Jeff Jarvis]

I was going to say a lot of. I mean, he, I assume, probably started off this trend of a lot of TechCrunch people going into venture capital or tech. Tech journalists have been doing that forever.

[Leo Laporte]

If you bought 1,000 bitcoin for $100, not even 20 years ago, 10 years ago, you'd be sitting pretty right now. You could afford a $60 million house in Miami.

[Paris Martineau]

Wow.

[Leo Laporte]

It's kind of amazing. Wow. Thank you for sharing that with me. Me?

[Paris Martineau]

Yeah.

[Leo Laporte]

Right. Well, that's it for this edition of Intelligent Machines. Jeff Jarvis, Paris Martineau. Wonderful to have you. Who are we talking to next week? Bonito. Do we have a.

[Benito]

Let me look. Yes, we have. We have somebody next week. It is Gary Rivlin, who wrote the Valley.

[Leo Laporte]

This is. He's got a new book about AI. The AI Valley. Awesome. You know Gary, right, Jeff?

[Paris Martineau]

No, I don't.

[Leo Laporte]

No, you don't. I thought you. I thought you must.

[Paris Martineau]

I don't think so.

[Leo Laporte]

He. He is. He's an investigative reporter. He's worked for the Times, Mother Jones, Wired. He's a Pulitzer Prize winner. So that'll be fun. He's been writing about Silicon Valley for some time.

[Paris Martineau]

New book, a lot of good books.

[Leo Laporte]

AI Valley, Microsoft, Google, and the trillion dollar race to cash in on artificial intelligence.

[Paris Martineau]

So it's focused on Google and AI and Microsoft.

[Leo Laporte]

Yeah.

[Paris Martineau]

That'll be interesting.

[Leo Laporte]

So it will be fun. That's next week. We are trying to shed light on AI, what it means, how you can use it, what it's going to be like in an AI driven future. And that's. That's all we do the week after that.

[Paris Martineau]

Can we talk? Can we mention that one, too?

[Leo Laporte]

Yeah. Who's the way after that.

[Paris Martineau]

You want to say it? Benito.

[Leo Laporte]

They've been doing a good job booking Mike McHugh.

[Paris Martineau]

Mike McHugh.

[Leo Laporte]

Oh, great.

[Paris Martineau]

So surf. We can hear about surf.

[Leo Laporte]

Yes. He's got a new venture. Mike is, of course, the founder of Flipboard. And I've been using Surf. The idea is it's kind of an aggregator. I guess you have been using it.

[Paris Martineau]

Oh, good.

[Leo Laporte]

Oh yeah, I have the, the test version of Surf on my, on my phone in Paris.

[Paris Martineau]

We're gonna miss you.

[Jeff Jarvis]

For the next two weeks. I am.

[Leo Laporte]

Paris will be in Paris next.

[Paris Martineau]

Now Paris. When I, when I went to Europe in the old days, I went ahead and did the show from Europe.

[Leo Laporte]

He did. Did. Sometimes from a basement, sometimes a cafeteria. You never knew where Jeff was going.

[Paris Martineau]

To show up because I, I. Paris was dedicated to the show and, and the cult.

[Jeff Jarvis]

That's true. You know, I'm not dedicated. I'm dedicated to eating a lot of cheese and drinking a lot of wine and not thinking about tech news for two weeks. I'm going to try and force myself to.

[Leo Laporte]

When do you leave?

[Jeff Jarvis]

Tuesday at midnight.

[Paris Martineau]

So in the, in the. Earlier on the Discord, I recommended a book. Book. I'm a huge fan of Russell Shorto and he wrote a book called Amsterdam, the World's Most. The History of the World's Most Liberal City. I think it is, ah, really good perspective on Amsterdam. Shorto also wrote.

[Leo Laporte]

I read his book on Manhattan, which I love.

[Paris Martineau]

Well, so he's got two. He just has a new one I just finished. You'll love it.

[Leo Laporte]

I read the island at the center.

[Paris Martineau]

Of the World, which is great. Which is about the Dutch.

[Leo Laporte]

New York, the early days of New York. It's fascinating.

[Paris Martineau]

I want to create a fund to build a statue to Adrian von der Dunk, who kind of brought the ideas of freedom of expression to our shores. He just has a new book out, which I just finished, called Taking Manhattan.

[Leo Laporte]

Oh, I'll have to get it.

[Paris Martineau]

Which is about how the English took over what was New Amsterdam and made it in New York and didn't fire a shot. Got and made it into kind of an experiment in a new economics and a new diversity and pluralism that becomes the essence of New York to this day.

[Leo Laporte]

And then along came Alexander Hamilton and the whole thing was shot.

[Paris Martineau]

There it goes. Shorter is just great. He's just wonderful.

[Leo Laporte]

And why are we mentioning him? You're gonna. We're gonna.

[Paris Martineau]

She's going to Amsterdam.

[Jeff Jarvis]

Oh, I'm going to Paris and Amsterdam. So I will.

[Leo Laporte]

So he's written a book about Amsterdam and New Amsterdam.

[Paris Martineau]

Yeah, he lived. He lived for some time. And so Amsterdam, A History of the World's Most Liberal City is really quite wonderful.

[Leo Laporte]

Now I've got to read the entire Russell Shorto library. The oeuvre. That's a French word that means eggs. Over easy, you know, O de fromage. Right.

[Paris Martineau]

How is your French, Paris?

[Jeff Jarvis]

Oh, it's a bit rusty. Is. I was. What I was going to say in French.

[Paris Martineau]

But are you going to. Are you going to impress the. The. The teammates?

[Jeff Jarvis]

I will definitely impress my skeeball teammates with my friends. I will never impress a French person. I did not do that even when I spoke fluent French in France.

[Leo Laporte]

The French people, though, it's interesting, like the American accent, because I always try to, you know, do all the R stuff and somebody told me, no, don't do that. Just do R. Because they love. They think the American accent is so cute. So if you say tres bien, they.

[Paris Martineau]

Go, ha, ha, ha, isn't she cute?

[Leo Laporte]

Instead of tres bien. Yeah, they'll. They'll just love it all.

[Jeff Jarvis]

I'll take that into consideration whenever I speak with. Or at least back in the day, whenever I was speaking in what I thought it was a French accent, one French shopkeep was like, oh, I thought you were in French. They're like, oh, I thought you were Belgian. You speak with a Belgian.

[Leo Laporte]

That's such an insult.

[Jeff Jarvis]

Because I. I know it's because my first French teacher was Belgian.

[Paris Martineau]

At least you weren't Quebecois. That would be an insult.

[Jeff Jarvis]

I mean, it's an insult, but it's not the worst.

[Leo Laporte]

Well, your French must be very good if you haven't. If they thought you were Belgian. That's very.

[Jeff Jarvis]

It was. It's. It's gotten worse.

[Leo Laporte]

It'll come, right?

[Jeff Jarvis]

I think it will. Yeah. I studied abroad there for a year.

[Leo Laporte]

Oh, my God.

[Paris Martineau]

Oh, wow.

[Leo Laporte]

In Belgium?

[Jeff Jarvis]

In France. In Paris.

[Leo Laporte]

Oh, you studied in Paris for a year.

[Paris Martineau]

Wow.

[Jeff Jarvis]

It was fun.

[Leo Laporte]

In high school or college?

[Jeff Jarvis]

In college.

[Paris Martineau]

Wow.

[Jeff Jarvis]

Partially.

[Paris Martineau]

Oh, yeah.

[Leo Laporte]

My daughter did that in her junior year in high school. She announced. She said, I'm going to be gone for a year, going to France. And it was. What? It was very disruptive. It was very disorienting. It was a sad day. She studied in Rennes for a year. She loved it. And we went and visited her, of course. It's beautiful. Well, have a wonderful trip. It's two weeks off.

[Paris Martineau]

We're gonna miss you.

[Leo Laporte]

We're replacing you with. With people who will be of a 10th year caliber.

[Paris Martineau]

Yes.

[Leo Laporte]

Mike, Aaron, I hope you tell them. Have a great trip. We'll see you in a couple of weeks. And thank you, all of you, for joining us. We do intelligent machines every Wednesday, 2pm Pacific, 5pm Eastern, 2100 UTC. You can watch us live on eight different streams if you're in the club.

[Anthony Aguirre]

Club.

[Leo Laporte]

Twit Discord's one of them. YouTube. YouTube is probably the preferred way. Most people like to watch it on YouTube, but there's also Twitch, TikTok, X.com, facebook, LinkedIn, and Kik. So choose your favorite platform. But you don't have to watch live. You know, we make it a podcast. You can download a copy of the show from our website, Twit TV IM. There's a YouTube channel dedicated to intelligent machines. You can watch the video there. Great way to share clips. Clips or subscribe. That's the best way, I think, in your favorite podcast client. And you'll get the show automatically the minute it's done. You have a choice of audio or video. I think most people now do video and then just listen to it if they're in the car. But if they want to see it, they can do video. But it's. It's up to you however you want to do it. We just have.

[Paris Martineau]

You want to see Leo shirts? You. You do.

[Leo Laporte]

The shirts are part of the attraction.

[Jeff Jarvis]

The shirts are a big deal. He's got a whole. How. How many shirts do you have? Have.

[Leo Laporte]

It's. I think it's approaching 50 now.

[Jeff Jarvis]

Whoa.

[Leo Laporte]

I think I. Look, I figured out that I could go three months without repeating. I think something like that. That's the goal.

[Jeff Jarvis]

Which one do you get the most compliments on when you wear it out? Is it still the avocado shirt?

[Leo Laporte]

They. Yeah, that's the one I wore in New York City. They love the avocado.

[Paris Martineau]

People on the street in New York who never talked to anybody, said, hey, avocado guy.

[Jeff Jarvis]

Hey, Yeah, I can't afford those anymore.

[Leo Laporte]

Avocados. This is a. Nobody's gonna say, hey, sugar skull guy. That's not gonna happen, so. But it is. This is a fun shirt. I agree. I get these from a little shop in San Miguel de Allende, Mexico, called Abrazos Design. You can get them online. They're wonderful. Made by local artisans. Local prince. Yeah, I'm really happy with them. They're very comp. Fair. Thanks, everybody. We'll see you next time on Intelligent Machines. Bon voyage, Paris.

[Paris Martineau]

Bo.

[Leo Laporte]

I'm not a human being.

[Jeff Jarvis]

Not into this animal scene.

[Anthony Aguirre]

I'm an intelligent machine.