D (3:46)

Well, I. I'm not talking about necessarily this particular decision. We don't know what the settlement's going to be.

A (3:54)

Right.

D (3:55)

I think it's a good thing to pay the authors when you use their book and training. And I have to say, I was on the fence about this, and this was a little awkward for me because I'm on the council of the Authors Guild.

A (4:12)

Ah.

D (4:13)

Which is suing these places.

A (4:16)

Right.

D (4:17)

They're involved in lawsuits. I have to say, I speak for myself, not the Authors Guild. And when we vote on these things, they poll the council and we vote. Should we sue and not sue? I abstain because I'm writing about these places. So I want to keep a little distance from going up to Sam Altman for an interview, and he's saying, by the way, aren't you suing me?

C (4:40)

I'll see you in court, Sam.

A (4:42)

Yeah.

D (4:42)

So it's basically the same as if Kevin Roose talks on the New York Times is suing him. You know, so I have a little arm's length of that. But I was, I think, a little unusual on the board. I'm a little more sympathetic to Silicon Valley than most of my colleagues on the board. And also I appreciate the counter argument to authors getting paid by saying that training a large language model model is something that takes in the world's knowledge, and it's kind of a good thing that we're able to put all that together and make use of it. Anyone could make use of it. So, on the other hand, you know, I'm a professional author, I'm a professional writer, I get paid, and it's not quite the same as, you know, a bright human being reading voraciously and then taking that work and using it in their work and their life, because this is something that millions of people are going to use, and that is a difference. But when I saw the anthropic settlement and it comes out to actually some. Not giant money, but real money, if it turns out that $3,000 a book, I've written eight books. My wife has written five. So, you know, that's a money that could help renovate a bathroom. Right.

A (6:07)

And that's $100,000 a book. That's a good deal. Yeah, yeah, yeah.

D (6:11)

So, and then I thought, you know what, I'm really okay with this. I really feel, because books are one of the most important corpuses that a large language model could use. Every book is a shot of a world. We build worlds when we write these books. And that is so valuable we, for the large language models to sort of understand humanity. And you know, it's very high quality content. It's, you know, original ideas. It's a lot of language, you know, spread out coherently. And I think those models will be impoverished without us. And now you look at the hundreds of billions of dollars that these companies are spending on infrastructure and suddenly that 1.5 billion, which to be sure is for pirating books, not for training them.

A (7:08)

Yeah. The judge ruled that they had fair use of the books they purchased.

C (7:12)

If they had just bought your books and you would have gotten two bucks.

A (7:15)

Each, then you wouldn't have gotten anything because they bought used books, by the way.

D (7:21)

So now, now I'm thinking, you know, this fair use interpretation, you know, is trying to apply in 1930s way of thinking to, you know, mid 21st century now doesn't work.

A (7:37)

Maybe it's changed since you wrote this article because since you, you wrote this judge also said, no, that's not a settlement. That's not a fair settlement. And in fact, the damages could run into the trillions for anthropic if they don't settle. Yeah. 150,000 per his complaint wasn't necessarily with the amount.

C (7:59)

It was with the structure in the lawyers.

A (8:01)

I understand. But by going back to the drawing board, there is this risk that it's going to be a hell of a lot more than 1 billion and a precedent.

D (8:10)

Yeah, but remember, that's for infringement. The, the precedent that really matters is the fair use.

A (8:15)

I agree.

D (8:16)

And this is step one on it. You know, there's going to be other steps and you know, what really there should be is a rethinking of, of the law. Though I'm a little worried under, considering who's running the country. That would work out for authors.

A (8:33)

But, but you would accept the fair use judgment or you prefer to get some payment from.

D (8:40)

I don't think that, you know, that interpretation of fair use in any way anticipated. AI.

B (8:47)

Yeah, you in the piece you Draw some parallels to kind of music licensing. Do you have a sense of like what a fair or workable kind of collective license system for books could even look?

D (8:59)

Yeah, I mean, I'm really glad you brought that up because one big reason, and Donald Trump cited this as the reason why you couldn't pay the authors for what their writings in a large language model would be because it's so hard to do. It's hard to figure out. But they figured it out without AI for music licenses. People go into bars and clubs and listen blog radio shows and they figure, figure out like what slices to give in streaming and all the rest of it. You know, I don't have the solution, but actually the Authors Guild is working with places to help figure that out, you know, because I think there, there can be some way to do that. You know, I'm talking to people now who are working in interpretability, in AI way, AI models, you know, to get inside the model and see how they work. And I think actually it would be quite feasible to have a model saying, you know what this output came because you know, it was 6% from hackers Stephen Levy's book. And then you know, there would be.

A (10:11)

With a link to the book itself so they could buy it.

D (10:15)

Yeah, it could be a link, you know, but, or, or it could be they a little like in Spotify, a little percentage of the pool, a fee to you.

C (10:24)

But, but that's part of the, part of the paradox of what's happening right now is companies like my former employer and now yours are cutting off the, with the exception of Google because they want to be in search. They're cutting off all of the AI companies. And Shipstead, the Norwegian company, just put out some stats today saying they've got a click through rate from ChatGPT of 1.2 to 5%. 5% by the way, for opinion pieces, not fact pieces, which is interesting. And so we're starting to see that AI is a new mechanism of discovery. So we kind of. Aren't we replaying the old search argument? How dare you take my soul by taking a picture of me. But I want the links from you. Are we going to end up in the same place with AI towards discovery?

D (11:07)

I don't think any links coming from.

C (11:10)

AI, but there is, there's data now and a place like Shipstead is making itself available and they're getting links and they're getting click through.

A (11:19)

But maybe another study books.

C (11:22)

Well, no, it's in fact the study that we talked about last week says that ChatGPT is not linking to social at all, but it is linking to editorial matter, which includes books.

B (11:32)

Yeah, but not book. Well, you're not going to link to a book because a book would be purchased.

C (11:37)

Well, that's part. That's to be the bigger question. Shouldn't we make our books linkable? It would be a good thing if they were.

A (11:43)

I guess you can always go to Stephen Levy dot com.

B (11:46)

That's true.

A (11:47)

And would that be acceptable to you, Stephen, instead of a micro payment that they, they say, you know, 13% of story came from hackers and then the link to Steven Levy dot com.

D (11:57)

No, they're getting money. They're getting money from that, from, from using.

A (12:01)

So you want a micropay.

D (12:02)

This is a, you know, it's, it's like one more, you know, you know, licensing opportunity. You know, just like someone might make a movie out of your book, you get some more money. Someone, you know, would run it as a first serial. You get some money. It's a, it's another way to get paid.

A (12:21)

I have heard the argument that there isn't that as much money as you think there is in AI. I mean, none of these companies are profitable.

D (12:31)

Well, right. Not right now, but they are going only under the assumption they think they will be trillions of dollars.

A (12:36)

Yeah, yeah, yeah.

D (12:38)

So if, if it doesn't happen, it doesn't happen. But if it does happen that a lot of that money is being made from using our material.

A (12:50)

I love your man from Michael Lunchbox behind you. By the way, I just saw Napoleon Solo peeking out.

C (12:55)

I like the Boss Gags album there. That's my favorite.

A (12:57)

And the Boss Gags album. I'm sorry, I'm getting distracted.

B (13:00)

Stepping back a minute. The three most recent pieces that you published, there's this thread of power consolidation among them. Do you think that AI is kind of accelerating this concentration of cultural and political power in tech or, or it's merely kind of just exposing what was already there?

D (13:20)

I think it's, you know, the former. I think it's accelerating it. You know, the. These right now is interesting, is a lot of companies, startups are trying to figure out how to, you know, find niches where, you know, they could plug into APIs, these models and have them be useful. But I think you're going to find that a lot of what they do will be fulfilled by the models themselves. So I think it is a centralizing movement and especially since the amount of money which is going to be involved in building out these infrastructures and they're going to need to pay that off is going to push them in that direction.

B (14:10)

I'm also curious, I mean, you've been writing about tech for decades. Does the AI era, like, feel like a genuinely new inflection point, or is it more of a chapter and evolution of kind of the same disruptive technologies clashing with social norms that you've been chronicling for ages?

D (14:27)

Well, I do feel that is one continuous story that, you know, like I am writing one story. Right. And I started when the PC era was booming. Leo can like this, right?

A (14:38)

I remember it well. Yes.

D (14:40)

Yeah. And, you know, and then on top of that, you know, came connectivity, came the Internet and then there was mobile and there was social and then, you know, and, you know, it, it and each thing built on top of the previous one. I think this does that. But I think some inflection points are bigger than others. The Internet was like a massive inflection point, you know, bigger than social, you know, I think bigger than mobile. But then, of course, that involved the connectivity. I think this is. It could be the biggest of all.

A (15:15)

Wow.

C (15:17)

Where do you come down on the doomsters on AGI? Do you believe AGI is actually going to. Is a legitimate goal?

A (15:26)

That is, you have Ray Kurzweil's book behind you. The Singularity is nearer. We had Ray on a few months ago. Yeah, yeah.

D (15:33)

No, I did an interview with him too. And, you know, he's really curious about what he thought it would be like to be like 400 years old, you.

B (15:42)

Know, and full of computronium. Yeah.

D (15:46)

Because you kind of keep listening to the music. You can see buzz gags you listen to when you were younger, you know, you know, you're like a teenager and, you know, when you're 400 years old, are you still going to listen to the Rolling Stones?

A (16:00)

Great question.

D (16:03)

Wow. And he said, no, no, you look for new music. I said, well, I'm not younger than 400 years old. And, you know, I. You look at my record collection, there's. There's not much stuff from this year, you know, but anyway, I think we have difficulty defining AGI. Right. You know, right now. I think for a little while it's going to be a moving goalpost, just like AI was. You know, I covered the Deep Blue match against Gary Kasparov and people thought, well, that answered things. And, you know, the Newsweek cover line for my story was the Brains Last Stand.

A (16:41)

Right.

D (16:42)

But it turned out.

B (16:45)

You had a lot of brain related covers, I suppose.

D (16:49)

Yeah. And then there's Einstein's Brain I was.

B (16:51)

Going to say famously found Einstein's brain.

D (16:53)

Yeah, but I, and I, and I just a couple of weeks ago I was, you know, talking to some of the people in the AI companies and, and they were saying, well, gee, now that we were saying we're almost in graphs of AGI, then we have to sort of think what AGI is. Right. You know, because the definition is meaningful, particularly it was meaningful between OpenAI and Microsoft, which had a clause about what happens when you reach AGI in their partnership, which in theory they've resolved, but they haven't shared how. But, but I do feel the thing is going to get better. I'm not like a Gary Marcus skeptic who's saying that, you know, this is, you know, a plateau and it's kind of a hoax. I do feel that things are going to get significantly better. I don't know what super intelligence really is supposed to mean, but I do believe that these things have a way to go to get better. And even if it doesn't happen all that fast, they're so good now that it's going to take a few years for us even to begin to exploit what we have in our hands now.

A (18:06)

I agree 100%. Yeah, but do you have concerns about it? I mean, you've been writing about these kinds of things going back to Artificial Life, the book you wrote some years ago. Do you have some concerns at all? It's not, not, you don't have to be a doomer, but absolutely, you know, Absolutely.

D (18:23)

I mean there's concerns of how, how well they do things right now. There's a number of things they do as well or better than people. There's a other category of things that they nailed pretty good, pretty good level, which is enough to knock out people's jobs. We've seen a couple studies recently which shows that in terms of employment, the people they're harming most, or entry level jobs, if companies are trying to save money by not hiring new people, because AI can do that, which is I think a big mistake because then where's their next level of executive going to come from if people can't start and learn and get better at it? But I think that there's a lot of problems we have from that it's going to automate probably already is malicious hacking and enable people to do a lot of bad stuff. And the temptation is just way too great for college students not to use that to prevent them from getting what they're paying for at these colleges, which is learning.

A (19:40)

Yeah, I have A lot of concern about government's use of AI as well. It really has become a tool for government oppression. But also I worry when you hear that, you know, the Social Security Administration is going to use AI to determine who gets benefits, it seems like a bad idea on the face of it. There are lots of misuses, I guess, of AI. As many poor ways to use it as there are good ways. Maybe more poor ways to use it than good ways. But you know what? You've covered technology since the dawn of the PC era. Every technology has pros and cons that come along with it. We lived in, right. We lived in an era where we thought the paperless office was just around the corner.

C (20:35)

Stephen, I'm really curious to get to your latest piece in Wired and your view of the culture of Silicon Valley, because you've covered the culture there more than really any other journalist. Did you say?

A (20:45)

I thought I knew. By the way, we should say just to reintro him. Stephen Levy, editor large at Wired magazine, Washington Post called him the. I forgot what it was. The premier technology journalist of our time. I think that's true. That's fair. I won't argue his piece from two days ago in Wired. I thought I knew Silicon Valley. I was wrong. Tech got what it wanted by electing Trump. A year later, it looks more like a suicide pact.

C (21:14)

Was it. Was it a change or was it the essence finally coming out?

D (21:19)

I somewhat more the latter. It certainly did not begin with. With Trump. So it had been, you know, over the last decade, certainly there a shift from, you know, where the, the ideals that I engaged with that drew me to this field were no longer so much in play, particularly among the bigger companies that I've been covering for so many years and the people running them.

A (21:50)

There was a book some years ago by John Markoff, what the Dormouse said, talking about how the 60s counterculture had influenced Silicon Valley. And you think of people like Steve Jobs as basically being a hippie, but there at some point was a shift in Silicon Valley towards aggressive capitalism, shall we say. And one of the things I think I've learned from the Trump era is that these guys never really had much of a commitment to the intersection of the arts and technology or DEI or any of this stuff, that these were merely pretexts to make more money. Do you agree?

D (22:34)

I don't know. From the very beginning that happened, I think there's a shift that happens. You know, I've known Mark Zuckerberg since 2006. I spent a lot of time with him when I did my book, and I think he was genuinely grappling with some of the effects of his product on people during that time. But at a certain point, you become so removed from everyone else. The money is so unbelievably huge.

A (23:07)

Yeah.

D (23:08)

You feel Marc Andreessen laid this out quite explicitly. You feel, wait a minute, I built this stuff, and here I am being whole before Congress. People are questioning me. I'm giving money to charity. You know, I've got foundations. Why don't people love me?

B (23:26)

Why don't people love the core of it?

D (23:29)

He called it the deal. He said, you make a lot of money, you build a company, you give money to give to good causes. And, you know, that is the deal. And the public and Congress and other people, they welched on the deal in his view.

A (23:46)

Huh?

D (23:47)

Yeah. Okay. And so it was so he's like, to hell with you.

A (23:52)

I remember Larry Page at a Google I O saying, we just wish there were a Google island where we didn't have to worry about regulations. You know, we could just do it. There is that sense. It went from being counterculture kind of hippies trying to save the world to we know better, and they should just let us let technology solve our problems. Yeah.

D (24:15)

Well, there was always a sense of, you know, we know better, but it was like, we know better to help everyone, you know. You know, a lot of times, you know, like Larry, he. Some things he did know better. I remember very early in the history of Google, he said, I don't. I'm not. They. They told him how much it would cost to install phone system. This is in the early 2000s. And he said, no, we don't. We're not going to install a system. Everyone will have cell phones. And people thought it was a scene.

A (24:46)

He was right. We haven't had a business phone system in a long time. Although tax law requires us to pay for people's cell phones if they use it for work. So.

C (25:01)

Do you have a landline, Stephen?

D (25:04)

No, I have. I keep my number with a. An Internet phone.

A (25:08)

Oh, yeah. Let's keep the old number. That's an old fashioned way of doing it. Right? Yeah.

C (25:13)

Void landline.

D (25:15)

Now they pull the car.

A (25:17)

Yeah. Really have it. Yeah. As they shut the copper. They don't want to. They don't want to maintain a copper infrastructure as value. What are you excited about for the next 10 years?

D (25:28)

You know, we're talking about the downside of AI, but I'm excited about the upside of AI. I'm excited to learn how. How I'm excited about how it learns. I'm excited about what it could teach us about intelligence. I'm excited to see what novel things these models can come up with to enrich humans.

A (25:53)

Yeah.

D (25:55)

Necessarily monetarily.

A (25:57)

Yeah. There's an opportunity and I think your pitch for equity for the content creators is fair. If there is going to be a lot of money to make in this, then. And we want AI to train on good quality material. I mean, I think it's.

B (26:13)

But it doesn't mean that it gets it all for free. Is that right?

A (26:15)

Sam Altman said you don't want to only train on stuff that we don't have to pay for because. Well, he didn't say pay for, but you don't want us to only pay for stuff that people give us because it won't be as good. You want to train on the best books in the world, you want to train on the best journalism in the world. But I think from a point of equity point of view, I think you make a good case that there should be some compensation. By the way, I love this picture from your website, stevenlevy.com of you. I guess it's a Vanity Fair shoot sitting next to Elizabeth Holmes of Theranos and a very young looking Mark Zuckerberg, Johnny Ives in the picture. A lot of people.

B (26:54)

The media power brokers cover or something like that.

A (26:57)

Is that Elon's head that's cut off at the top there? I don't.

B (27:01)

Elon was there for sure.

D (27:04)

Standing up.

A (27:05)

Yeah. So I was wondering who that was. So that's Elon being cut off there.

D (27:09)

Yeah, yeah. And you know, there's Lena Dunham, you know, like my name.

B (27:14)

Oh, that's Jessica Lessons over there in the top, right.

A (27:17)

Oh, yeah, there's Jessica, your old boss. Yeah.

D (27:20)

They somehow let the moderators into the photo shoot. It was one of those classic Annie Leibovitz shoots where.

B (27:26)

Oh, this looks very Annie Leibovich.

D (27:29)

Yeah. So we, we went off the stage and I was interviewing Zuckerberg and Michael Abrash, who's the head of his AI lab. VR lab at the time, not AI. And you know, so the three of us were put into a room. There were people sitting in the room, posed the way we were going to be sitting.

A (27:48)

Right.

D (27:48)

And she had us assume those poses.

A (27:52)

Here's your stand in, Stephen. You sit there like that person.

D (27:56)

Yeah. And then the next panel came in after they were done and then she had people standing there. So we all look like we're together. We were.

A (28:05)

Oh, they cut it in oh yeah, isn't that. Yeah. Maria Shriver. It's an interesting group.

C (28:09)

You can do that with, with Google now. But yeah, yeah.

A (28:15)

One last, one last question. Since you had, you know, you do have the kind of. The book about Facebook. What do you think of Facebook's pivot from VR, the Metaverse? They even changed their name to now AI. And these latest glasses, the Ray Ban.

D (28:32)

Displays, well, you could blame mobile for that. When mobile came about, and this is something I found in the book, Facebook had a near death experience. They were ahead of everyone else because they were the first really big company that was totally on a web infrastructure. All the other companies were updating every six months and they update every three hours. Maybe you know that, you know, because it was a totally web based, you know, infrastructure. Then mobile came along and it was a challenge for them because people had phones to be updated and they, they didn't know how to deal with that. They didn't know how to do ads on it. And they went on a crash course to learn and figure out how to do it. And they survived. But he told himself, I'm never going to be in that position again. The next paradigm that comes on along, I'm going to be on top of. He identified it as VR, which isn't quite happen. Doesn't quite happen. Yeah, yeah. You know, he said, In 10, 10 years after I buy Oculus, that's going to happen. And it's more than 10 years and it didn't happen. But he managed to salvage something out of it with the wearables, with the glasses. And then he's like, oh my God, it's AI. Everyone's ahead of me on AI. So he's throwing huge amounts of money to be relevant in AI and now they're going to try to build it into those glasses. But you know, when he senses, his Spidey sense goes up that saying there's a paradigm coming up that could unseat us because that's when companies with a monopoly on something, you know, get overthrown. When a new paradigm comes on and everyone, you know, leaves an app and goes on to something power by the next thing you know, he wants to react to that threat and be ahead of the game. So that's what's driving Hard to do. It is hard to do.

A (30:26)

Yeah.

D (30:27)

You know, Apple, Apple did it.

A (30:29)

Yeah, Apple did it. We don't know if they're going to continue to do it. That seems to be the question mark.

D (30:35)

Yeah, yeah, we don't know where they are on AI.

A (30:37)

Yep. Stephen Such a pleasure. America's premier technology journalist. Read every one of his books. And are you working on a new one?

D (30:48)

Not at the moment. I'm zero in on one.

A (30:50)

They keep me busy. Yeah, Wired. You're writing a lot for Wired. He's editor at large at Wired. And that's a great place to read Stephen's works as well. Stephen, thank you so much for spending some time.

D (31:02)

It's always great. And I love you and your fellow panelists there.

A (31:06)

Yeah. Well, we'll get you back anytime. Thank you, Stephen. All right, take care, Steve, everybody. Hi. America's premier technology journalist, an inspiration to us all. Absolutely. We'll take a little break and then we come back. We've got AI News. How about that? You're watching Intelligent Machines with Paris Martineau and Jeff Jarvis. I've been feeling fit as a fiddle lately, and I think it's because of our brand new sponsor this week, Field of Greens. And it's, you know, it's not new to me. I've been consuming this superfood for some time and it's helped me sleep better, it's helped me feel better. And apparently I did not know this until I talked to the folks, the doctor behind Brickhouse Nutrition, the creators of Field of Greens. Apparently, it's made me younger. Think of it as a, well, you know, a hard reset on a computer, put you back at the factory settings. Everything seems to run better and faster. Right. This is a smart reset of your health to get you back to, if you will, the factory, the factory settings. And they know it works. This is an amazing. In fact, you know what? I'm going to mix some up right now. I have the strawberry lemonade flavor, although they have a variety of wonderful flavors. And I'm going to drink some just to prove to you not only that I drink it regularly, but that I love the flavor. Oh, that's so good. They did a study with Auburn University for a pretty sizable biological age study, and they did it right. They took a large cohort of people, did a lot of blood work on them. Did, you know, a complete workup on them to see where their chronological age and their physical age paired up. And some people were not in the greatest shape. You know, they were kind of physically older than they should be for their chronological age. They gave these people, some of them placebos and some of them the Field of Greens over a period of time. Inside Field of Greens, fruits, vegetables medically selected for specific health benefits. In fact, you go to the website field of greens.com and take a look at the label. It's, it's got everything. I mean, all the good stuff is in here. And it's, it's targeted at different parts. So there's a part of it targeted to your blood work, some of it's targeted toward your biometrics. There's a, there's a mineral blend in here. So they gave this field of greens to half the population in this medical study at Auburn. You can read the whole study on the website. They gave the rest, placebos, measurable, measurable health benefits to the group that did the study and drank the fields of green. Considerably more than the placebo group. They were told not to change anything, by the way. They didn't tell the group, change how you eat. Nope. Eat normally. If you drink, fine. If you don't, exercise, fine. They wanted to eliminate all the variables so that they could see if field of greens changed anything. And the results, frankly, quite remarkable. The group that added field of greens without changing any of their lifestyle literally slowed how fast their bodies were aging. They changed their physical age. I can even see this on my Oura ring and my apple watch and my why things scale. My, my biological age is. It's not aging as fast as my chronological age. The field of greens, all organic. They have a variety, as I said, of ingredients in here for heart health, there's a group of things. For cells, lungs, kidneys, liver health, there's a metabolism group for a healthy weight. It helps me sleep better. It's really an amazing treat, frankly. So I want you to go to the website, check out the university study, and you'll get 20% off if you use our promo code, IM. Go to fieldofgreens.com fieldofgreens.com promo code is IM. It's very refreshing, very delicious. It mixes up very easily. 30 calories, by the way. So it's low calorie. It's the equivalent of 2 servings of fruits and vegetables. And the whole family can drink it. There's antioxidants in here. It boosts your immunity. Fieldofgreens.com from Brickhouse Nutrition. We thank them so much for the support. And if you want guys, I will be glad to send you some field of greens. You can try for yourself. Jeff, you need field of greens. I think you're chronological age.

C (36:03)

I need to slow down, put the.

A (36:05)

Brakes on who does it at our age, you know? Yeah, it's good. Let's take a look at the AI news. Paris Martineau. Jeff Jarvis. There is so much AI news. I'm kind of curious what you so much AI news. You know, when we started this, there was. I thought, are we going to have enough to talk about every week? Holy cow. One of the things I thought was kind of interesting, there have been a spate of investments.

B (36:39)

Yeah. Can we talk about the circular Nvidia opening?

A (36:42)

Okay, I wasn't going to say circular, but you said the word the correct.

B (36:46)

Way to describe it.

C (36:48)

I think it is so money to give to us.

A (36:51)

Nvidia gave open a $100 billion. It's actually spread out over 10 years. But that's a lot of money to buy $100 billion worth of Nvidia chips, so. And by the way, we've been seeing this a lot this year with companies this kind of circular stuff. And I think there are some who say Financial Times says this is essentially a shell game, that there is no net gain or loss for either company. But it looks good on your balance sheet.

C (37:23)

You can put it in your balance sheet.

B (37:24)

The Wall Street Journal broke it down like this. OpenAI will use the cash from Nvidia's investments to for new chips produced by Nvidia, a circular arrangement that allows the chip company to turn its balance sheet cash into new revenue. Such circular arrangements are common in the AI world. Wall Street Journal writes and have raised questions about the extent to which new sales in this industry reflect genuine market demand versus capital recycled within the industry itself. Which I think is where it all that is what it all boils down to. It's like, is this another major foot in the gas that is fueling kind of this hype cyc and how do we think about this industry in this market when you have all these different kind of properties boosting it up like what is. How do we separate the real from the fake here? And it gets difficult when you have things like this.

A (38:17)

It's not just Nvidia and OpenAI. Oracle did exactly the same thing and.

B (38:23)

They got the biggest stock boom they've seen in like three decades.

A (38:26)

So it paid for itself, right?

C (38:28)

Oh, yeah. Well, it allowed the Ellisons to then turn around. According to Oliver Darcy's media podcast, the increase in Oracle stock was double the price they're going to pay to get Warner Brothers Discovery and thus cnn.

A (38:45)

So the network is free.

C (38:49)

Messing with democracy is free.

A (38:51)

Oh, man. OpenAI has teamed with Oracle and SoftBank to build five, count them, five new Stargate data centers. Seven gigawatts of power.

C (39:02)

I got a quick question for you there, Leo. I can't remember who asked it on the socials this week. But somebody said, when did we start measuring compute in gigawatts?

A (39:10)

When it started using so much power, I guess because it isn't a normal way to measure it for sure.

C (39:16)

How would. Was it flops? How would you measure?

A (39:20)

Yeah. In fact, for AI, we often use tops, which is trillions of operations per second. But I guess these network operations centers, these AI centers, these. They're all big GPU farms, basically. Yeah. And maybe it's easier to measure it in the massive amount of power it's using.

C (39:39)

Did I, did I give you the number for the, for the calculator? For the Texas OpenAI that they're building, they're going to have. If you look at how many transistors, there's 280 billion transistors on a blackmail chip.

A (39:53)

Yeah.

C (39:53)

You're going to have 112, quadrillion transistors.

A (39:57)

See, this is why you say instead 4.5 gigawatts. That's the Abilene Stargate facility they're already building. These are new sites. Shackelford County, Texas. Dona Ana County, New Mexico.

C (40:10)

Here's the next question. They need water. Why are they being put in dry places?

A (40:15)

That's a very simple answer to that. Tax breaks.

C (40:17)

Okay.

A (40:18)

And it's a shameful thing. I mean, the state of Arizona is giving these companies massive tax breaks, but they are basically, how can I say this, Borrowing their water from the Colorado river and depriving people upstream and downstream, I should say, of the water and the power. I mean, this is crazy. Some of this is driven by the administration, which likes those big numbers. Remember when Mark Zuckerberg leaned in to the president and said, 600 billion. Is that the number you wanted me to say? He likes those big, round, large numbers. And of course, the President did in fact announce the initiative, the Stargate initiative, the day he walked into the White House. Stargate will be. Is an umbrella name. It's kind of unclear what Stargate is used to refer to all of OpenAI's data center projects except those developed with Microsoft. So the Abilene site is owned and operated by oracle, I think. $30 billion.

B (41:28)

I was going to say. Aren't there some of the things that fall into the Stargate label that have been announced but have yet to materialize in any real form? That's different than this, what we just discussed? Yes, those ones are materializing.

A (41:42)

Yes.

C (41:43)

Okay, well, they say they are materializing.

B (41:47)

Question mark, Asterisk.

C (41:48)

Yeah, right.

A (41:49)

Yeah.

C (41:50)

Parenthesis.

B (41:51)

Parenthesis.

A (41:53)

Sam Altman said something about gigawatts today. What was it? Like, what did he say it was? He said our vision is simple. This is from his piece on a. Oh, I just closed it. Darn it. There's a bug in the comment browser that when I try to zoom in on text.

C (42:11)

You're using Comet.

A (42:12)

Well, I thought I'd try it. There's a bug, though, when I zoom in on text, it crashes it, so I have to use the. Yeah, well. And I zoom in on text all the time for the show because it's easier to read. So this is the. The blog post from yesterday from Sam Altman titled Abundant Intelligence.

B (42:31)

Oh, no.

A (42:33)

He says that's just a combination of.

B (42:35)

Buzzwords I don't need.

A (42:36)

If AI stays on the trajectory, we think it will, by the way.

B (42:41)

Okay, already off to a crazy start.

A (42:43)

Yeah, right. We even kind of referred to this with Stephen, that we're kind of plateauing, I think, a little bit. And the expectations for AI seem to have been plat anyway, but not in Sam Altman's mind. He says amazing things will be possible. Maybe. And this is the number I was looking for. With 10 gigawatts of computer, AI can figure out how to cure cancer. Maybe.

B (43:08)

Maybe if we actually focus any of the efforts on it. But instead, that's not going to make us a lot of money right now. Instead, we're going to figure out how to get the chat bot to be.

A (43:17)

Your therapist, your sex buddy, or with 10 gigawatts. I know, and I know I went there. With 10 gigawatts of compute, AI, Sam writes, can figure out how to provide customized tutoring to every student on Earth. He says if we're. So that's one, that's another 10. So we got 10 and 10, that's 20 total. He says if we're limited by compute, we're going to have to choose which one do you want to solve cancer or ignorance, or do you want to tutor every student on Earth? No one wants to make that choice. No one wants to make that choice. So let's go build.

E (43:53)

Actually, I think we all want to cure cancer. I think that's the choice we would all take.

A (43:57)

Cure cancer. Really, Sam, that is probably best.

C (44:00)

Somebody who's had it twice.

A (44:01)

Yeah, yeah. I think we're all in cancer's waiting room just waiting for the doctor to come to the door and say, leo. Is there a Leo here? This is the clincher. Our vision is simple, says Sam Altman. We want to create a factory. A factory that can produce a gigawatt of new AI infrastructure every week. Every week. Jesus. The execution of this will be extremely difficult.

C (44:35)

We don't have lots of money.

A (44:36)

It will take us years to get to this milestone. It will require innovation at every level of the stack, from chips to power to building to robotics. But we have been hard at work on this and we believe it's possible. It will be the coolest and most important infrastructure product ever. It'll be cool. Cool. We're particularly excited to build this in the us Although. Oh, let's point out, a lot of this is going to happen in the Middle East.

C (45:02)

Yeah. They also don't have a lot of water.

A (45:05)

Yeah.

B (45:06)

Why? Listen, I know we've talked about this before, but why do they keep building the things that need to be cold in the hot places? I know it's because. I know logically it's because tax breaks and they have a lot of land where people don't want to live, but it's like, come on guys, why don't go.

A (45:24)

Now we know the strategy. Oracle. Oracle's deal with OpenAI is for 10 gigawatts of data center capacity. Nvidia is putting 100 billion in OpenAI. The initial build out will be 4 to 5 million Nvidia GPUs. The chip maker will make those investments though, spread out progressively as each gigawatt is deployed. We're talking more like a gigawatt a year than a gigawatt a week.

C (45:55)

But.

A (45:55)

So.

C (45:56)

So, Leo, I've been thinking about this. We talk about the scale for scale's sake thing a lot. And, and I had this discussion with Jason earlier today. I'm studying right now, researching the early history of radio. And when Marconi was sent, it was, was doing sparks that could then be detected elsewhere. The sparks were, were damped signals. They, they went out on every, every, every frequency there was. There was no way to tune. They didn't know how to tune, they didn't know synthete yet. So the only way that they could go farther was more power and so, and bigger antennas.

A (46:30)

It's funny, my radio era, you would always talk about how many watts that's true. The radio station could transmit it. Yeah. When I worked at the 50,000 watt clear channel KFI in Los Angeles, that's as good as you could get. 50,000 watts.

C (46:47)

Right.

A (46:48)

And Clear Channel meant that there was no one else on that frequency anywhere in the United States, which means you own that frequency. And theoretically, if the propagation is correct, you could, you could, you could bounce all over the world. Yeah.

C (47:00)

So, so they bought ever bigger dynamos, ever bigger power, until they learned how to tune and then it became a whole different thing. And it just strikes me that, that these guys don't know how their thing operates. So the only thing they know to do is to throw more power. More power, more power because they don't know how to tune it. They don't know how to that next level of making it work in a different way. And so we're going to throw, as we talked about with Karen Howe, we're going to throw just ungodly amounts of money behind one philosophy, which is more gigawatts, more gigawatts, more gigawatts. Surely once we do that, we'll cure cancer. We don't know how, but it'll just happen.

A (47:39)

Here's the counterargument. Please, we don't know if it won't work.

C (47:46)

Go ahead, Paris.

B (47:46)

Yeah, but there's also no, like, okay.

A (47:51)

One, there's evidence that it does work. Wait a minute.

B (47:53)

There is evidence that it does as an argument for this, but also it's worked so far. Weird. No, we're doing, are not, we're not investing all of this time, money and resources into AI powered endeavors to cure cancer. We're investing all this time, money and resources primarily into commercial chat bots and enterprise products and siphoning, as we talked about with Karen Howe, siphoning resources and brainpower away from the fields of AI research that are most applicable to medical breakthroughs.

A (48:29)

Well, so you say.

E (48:32)

I keep, I keep saying this but like if, if open AI cures cancer, they will have the goodwill of the people forever and they could do whatever they want.

A (48:40)

And that's why the VCs are investing. You know, there is this and, and, and to be fair, there's sort of evidence that throwing more, I mean, that's what's gone on so far.

B (48:54)

There's evidence that what throwing more.

A (48:57)

So far, throwing more compute at AI.

C (48:59)

This goes, this goes to the, the FT says that companies keep talking about AI but they can't really explain the upsides, that it can't really get to the results.

A (49:08)

We'll know when we get there. For crying out loud. You guys act like we're driving to Las Vegas.

B (49:15)

Like there should be some more reasoning for giving like a handful of.

A (49:20)

Here's the reason.

B (49:20)

All the money in the world.

A (49:22)

Here's the reasoning. We were at chat GB to 2.5, if you remember using that garbage. Hot trash. It just was nuts. Then all of a sudden, by making bigger models, we got to ChatGPT 3.5 and it was suddenly A quantum leap in the capabilities of the AI. And that was when OpenAI and Karen Howe talks about this, came up with their, you know, the law, OpenAI Law of Scaling, which is, you know, throw more compute at it. Remember the famous, what was it? The bitter lesson article that we talked about which said basically you don't have to make better models, you just have to throw more computer.

C (50:08)

You also have. We talked about Yann Lecun last week saying stop with the ll says that.

A (50:12)

But the evidence is not in his favor. This is his thought. But the, but the evidence so far, how long it took.

C (50:19)

I also just studied how long it took to invent the transistor and they focused on it and they didn't say we have the first clumsy transistor, we're going to make a billion of them. No, they made it better. They got to a point where they knew that it was scalable and they knew that it was something worth doing. I'm going to do the thing you hate and ask you to go to a line, line 30 yet?

A (50:40)

Not yet.

C (50:41)

Well, okay, but the point is that there's, there's economic money here that says that they're going to be, According to Bain, $800 billion revenue shortfall in this all made up.

A (50:51)

We don't know. We don't know. I'll give you the point. I am arguing this a little more strongly than I actually believe, but I, but I think it's important to make.

C (51:00)

This argument for the warning.

A (51:02)

All of you know we don't know what's. We don't know. Yann Lecun doesn't know. No one doesn't know. But, but the evidence has been that the. We keep building these bigger models and they keep getting better. Now Martin Piers in the Information yesterday wrote a piece responding to Sam Altman which is can we afford AI? Which addresses this exact issue. And he says countless news reports cite businesses who say they've yet to see enough benefit in AI to pay for it. The Information reported last week Microsoft having a hard time persuading its customers that AI features in Office were worth the price. And he also cites that Financial Times argument article that says big companies find it difficult to explain AI's value. I don't think though that that, I mean, let's not be nervous. Nellie's here with lack of resolve guarantees failure. Why don't we throw all our resources at this? All the evidence is it works.

C (52:11)

No, no, no. Define works. Define works.

A (52:14)

Compare chat TB2.5 to the latest models and I think it works pretty well.

C (52:22)

Works for madness, works at economic value. That's the, that's what's missing here. We see a lot. I was on a World Economic Forum. I don't get invited to Davos anymore, but I get on the, on the, on the zoom calls. So I was on a call this week.

A (52:37)

Come to Switzerland, but you can call us.

C (52:39)

Yes. So I was on a call this week and one of the executives there said, oh, we're seeing real business efforts, business benefit. I said, okay, I'm really eager to hear more. And I said to the forum, you should be gathering those receipts because I pointed out what Bain said and what the FT said and what the Deutsche bank just said. And I think there's a real crash of reality hitting unless we start to see arguments that include receipts for what the economic benefits are. So far, I'm not saying Leo is going to be perfect.

A (53:12)

I'm not saying you're like everything the people who said when John F. Kennedy said, In 10 years, before this decade is out, we will put a man on the moon. You're like the people said, yeah, but why? What's the benefit?

B (53:22)

Tang I. Tang, Tang.

C (53:26)

We got Tang out of it.

A (53:27)

Beating the Russians. Ballpoint pens that would write in upside down.

E (53:32)

And the Cold War.

B (53:35)

Yeah. What's the downfall?

C (53:36)

And the end of humanities and universities and everything became stem.

B (53:42)

Mit. There was a really interesting study out of MIT this week that you probably saw picked up because there was a Harvard Business Review piece highlighting a takeaway from it which was related to work.

C (53:53)

Slop, which is a great word, by the way.

B (53:55)

It's great. They basically they reviewed like hundreds of publicly disclosed AI initiatives, interviewed representatives from 52 organizations, interviewed hundreds of senior leaders across AI companies and others and found that that Jenny is flooding a lot of companies that use it with low value work slop. Which means like I generated drafts, decks, tickets, notes. So output ostensibly explodes, but business value doesn't because despite widespread adoption, a lot of the firms are not seeing measurable ROI improvements from Jenny I because they're scaling activity, but the outcomes are not scaling the same pace because people have to spend a lot of time double and triple checking and correcting these and kind of it's. I think this is part of what we need to think about when we think about what AI is providing to our society. Because right now a lot of the most obvious benefits are a bit more complicated.

A (54:53)

I think you are nattering nabobs of negativism, which.

B (54:58)

No, sir.

A (55:00)

Oh yes. I thought you'd recognize what happened to the vice President who said that, you know, sometimes people could take $50,000 bags of cash and get in trouble, sometimes not. And it really. It just depends. It just depends on the times. On the times. My point is what if. What if, you know, it's true that scaling produces results. The evidence is it has been producing results. And we say, oh, yeah, well, never mind. And we miss out on that cure for cancer which could be just around the corner.

B (55:39)

See that 0% chance that at this current stage that we're going to stop invest. We, the royal. We are going to stop investing money into AI that is happening.

A (55:49)

So your defense is my argument will fail.

C (55:52)

So are there other investments that may be wiser as well?

B (55:56)

At the same time, I just think that my argument for this has always been no matter what technology you're talking about, you could actually have computer. Only about their development. And I. My argument would be it seems unwise to bet everything on one very.

A (56:13)

Are we betting everything, though? Are we really betting?

B (56:16)

We're betting the vast, vast majority of venture capital and like just general resources available in our society. Yeah.

A (56:26)

Alibaba. The new QN models are remarkable models, especially for AI coding. And these are, to my knowledge, models that are made much less expensively.

C (56:38)

That's interesting. Yeah, that's interesting.

A (56:41)

I think there are. I don't think. Look, there's opportunity in a variety of areas and I think there are certainly people who say, you know, what if could we do this without a gigawatt a week of power? And I think those people are still working.

C (56:56)

Leo, there's a chart at line 114 I would love for you to share.

A (56:58)

Is this the one you wanted before? I don't.

C (57:00)

This is bake. Yeah. And. And it's.

A (57:02)

You trust Bain.

C (57:04)

No. Well, they're a little better. McKenzie.

A (57:07)

That's not saying anything.

C (57:08)

No, I know, but, but it's. But it's. This is one interesting that I think that really will strike. They. They looked at the training.

A (57:14)

Compute in flops, floating operations per second. Okay.

C (57:18)

Yeah. Compute demand growth is more than double Moore's Law, which I found fascinating.

A (57:26)

That was kind of the OpenAI insight is that that was their new law replaced Moore's Law. Yeah.

C (57:30)

Well, but, but it's demand. It's not supply, it's demand. So compute demand growth is four and a half times every year, whereas Moore's Law is two times every two years. So, you know, it used to be we got better radios and better computers and better cars and better everything because of Moore's Law.

A (57:48)

Right.

C (57:48)

The higher Compute at the same or lower cost gave us more functionality. That's what's killed here is not because Moore's Law is dead. It's because they're outpacing it so much. Their hunger for the compute for the gigawatts is so huge. And that's why Bain said we're going to be in an $800 billion shortfall in 2030. And so if you're going to be, and if it's worth the investment to society, okay, but by then you better have some more receipts and justification for that investment. I would say.

A (58:25)

Well, I would say that people are spending the money, get to decide whether they feel like it's worth it. Right. It's not my money.

C (58:32)

In an oligarchy, it does. But maybe if we had a fair taxation system. But that's a different topic.

E (58:39)

There's also one maybe.

C (58:40)

Maybe if we were.

E (58:43)

New technologies almost always first are used as weapons. So, I mean, there's that really.

A (58:50)

All right, so wait, wait, wait, wait.

C (58:52)

NATO, Really?

A (58:54)

I think military weapons that can do.

E (58:56)

Things, that can be weaponized research is for sure.

A (58:58)

Often the Internet came out of military research, I think military research.

C (59:02)

The Navy had a big role in radio.

A (59:03)

Yeah, okay, atomic bomb. But that's because. I'll tell you why that is. That's because governments tend to invest in defense.

E (59:12)

Yes. Hence, it's the most profitable branch of this.

A (59:14)

And government has more money, or until recently, has had more money to invest in R D and pure R and D than anybody else. So the R D budget has gone towards the weapons of mass.

C (59:26)

Well, they still have money to spend on weapons of defense. They're just not spending it. They're unwisely spending.

A (59:30)

I'm sorry, wouldn't you prefer that they spent the money on AI instead of bigger bombs?

C (59:38)

More drones, Leo, Forget the bigger bots. Now it's all drones. Drones. Drones.

A (59:41)

What's one word?

E (59:42)

Palantir.

A (59:45)

Well, yeah, as you know, I have definitely qualms about, hey, AI is used. Being used. I just saw that Llama Meta's AI system has been approved for use by federal agencies. What could possibly, possibly go wrong? This is not, by the way, far from the best AI model out there. But as you know, Google and OpenAI have been giving their AIs to government for a buck a year. Both of them better than Grok. I'm sure Grok's next.

C (60:17)

Why isn't it already there? I thought he already did a deal.

A (60:19)

Oh yeah, Doge with Grok. Yeah, maybe. Yeah. The GSA has also signed by the way, the gsa, which is the government agency that has been so depleted by DOGE that they're now trying to rehire some of those employees. They signed off in recent months on AI tools from Amazon Web Services, Microsoft, Google, Anthropic and OpenAI. All of those companies agreeing to sell their products at steep discounts and to meet the government's security requirements. So LLAMA is just the latest in the batch. I think that you're right. There's some places I don't want to see AI used, but I really think it's worth spending the money if we've got it. If we don't have it, we don't have it. That was. The Bain study was saying it's going to cost half a trillion dollars a year and we don't have it. And if that's the case, then it won't happen.

C (61:14)

The other thing that's interesting is one of these stories said that basically all of this spending is what's preventing us from right now being in a recession. Well, we're at a teetering edge that if these stories like the Bain and the Deutsche bank and the FT stories start to stop.

A (61:32)

Stop peeing in my Cheerios, guys. I got a 401k to maintain here.

C (61:38)

Same here.

A (61:42)

I really. I often wonder whether I should just get out of the stock market and just hold gold. But then I think, well.

B (61:53)

Gotta invest.

A (61:54)

I don't know what I think. I think, what the hell.

C (61:57)

So I have. My wife gets nervous about cash and I cashed out some stuff. So I cashed out. I've got one mutual fund that's obviously filled with tech stuff because it's the hot stuff. And, and I took some off the table and my broker put it in a bond where I'm making some decent returns.

A (62:15)

Less than inflation, but okay, but, but.

C (62:18)

Then I looked at what I would have made in that mutual fund.

A (62:22)

Oh, don't do that. Yeah, yeah, no, I've stayed. I don't, I don't. I only own index funds, but I've stayed in the index funds because you know what? If I have to work until I'm 104, so be it. I like my job, but, you know.

C (62:39)

What episode number will we be that?

A (62:42)

Ask AI. China's Deep Seek says that its best AI model costs $294,000 to train.

C (62:53)

This is where I have some hope.

A (62:56)

Yeah, or they're lying.

C (62:58)

Yeah, that's quite possible.

A (62:59)

That's the R1.

B (63:00)

Everyone could always be lying. I think this is something really important. To consider.

C (63:04)

Yes.

A (63:06)

This is in a peer reviewed article in Nature published on Wednesday.

C (63:11)

There you go.

A (63:12)

Deep seeks release of what it said were lower cost AI systems in January caused a, as you may remember, a plummet in the valuation of Nvidia and other tech stocks. They've since recovered, of course. The Nature article, which listed the company founder Liangwenfeng as one of the co authors, said Deepseek's reasoning focused R1 model used 512 Nvidia H800 chips. This is new information. We didn't have this before and it cost, as I said, less than $300,000 to train. Meanwhile, Sam Altman said that his foundation models cost much more than $100 million each. Now, incidentally, China has decided not to buy any more chips from Nvidia. Yeah, so that's over. The real question is whether Hyundai and other companies that are selling to China can develop chips to compete. So far, no. I, you know, I think it would be a mistake for us to push the pause button at this point.

C (64:21)

I'm not pushing the pause at all. I'm pushing just a different mix, which I think Paris, you would say too well.

A (64:27)

But who are you pushing that to? The people who are spending the money.

C (64:29)

Research. Well, I would like to. We've, we've, we're shutting off research. That's what disturbs me most. Because God knows what. We don't know yet.

A (64:39)

Yeah. Taking a break. More to come. You're watching intelligent machines or listening. Either way, there's not much to see here except my lovely rainbow colored. It is very nice shirt. This is a very colorful studio. Is it, is it, is it gaudy, do you think? Is it gauche? Am I. Should I find more?

B (65:03)

Dude, it's not.

C (65:05)

No.

A (65:06)

I feel like I'm competing with all those YouTubers. I gotta, I gotta give you something.

B (65:11)

You need to have a setup. Kind of like the people do on those tick tock lives where it's like a woman sleeping and if you donate some coins to the chat, it like drops an object on them while they're sleeping. Live. You kind of incorporate that where people could pay you five or ten dollars to like slime you or something.

A (65:29)

I know I shouldn't probably tell you this, but your predecessor on this show, Stacy Higginbotham, had a design for a punching bag glove that would hit me in the head if people paid enough money.

B (65:41)

Oh, I think that's really a good idea.

A (65:43)

That one or Stacy, address the primary.

B (65:47)

Issue issue of sliming you, which is that it would mess up your Makeup, so, you know, it could just give it a little sock to the face.

A (65:52)

I powder my nose a little bit more here.

E (65:55)

I mean, you can absolutely monetize the boxes above on your. On your left and right. That's like.

B (66:03)

Yeah, someone.

A (66:06)

Learned. It's boxing the ears. You go like this to the guy's ears. Tai Chi is 20 bucks to put.

B (66:11)

I'm with. I'm with idiot. And then like, have like a little, like one of those T shirts with a little arrow.

A (66:20)

I keep having to turn off the Apple camera reactions on this machine because it's.

B (66:26)

I've never had mine turned on. I don't know how to do it.

A (66:29)

Mine keeps turning on for some reason, which is extremely.

C (66:33)

For me. Zoom just does it. I guess. It's not doing it right now.

E (66:37)

It depends how we're producing the show. If we're doing it through Restream, it's not actually your computer that's doing it. It's the computer in the cloud that's.

A (66:43)

Doing it when you do it.

C (66:44)

You see?

A (66:45)

Oh, okay. Yeah. Because I was having that problem yesterday, and I was very concerned about it.

E (66:52)

So we need to turn it off in the cloud. That's what we need to do.

A (66:55)

Turn it off in the cloud. Okay. By the way, Poco with a k has asked AI and our Intelligent Machine show, when I am 103, which will be September 24, 2063, will be episode 2737.

C (67:12)

I'm ready Paris be.

A (67:13)

Then if I keep drinking the field of greens, I might actually make it. I might actually make. Is my goal to keep doing this as long as I can. Now Lisa wants me to do a daily news show.

C (67:29)

That again.

A (67:30)

Well, what that means.

B (67:31)

Are you gonna do TI tv?

A (67:33)

It'll be. No, not titv. No. We'll leave that to the information we're thinking of. We have the name Tech News Today, which is tnt. And I kind of like that, so we'll probably keep that.

C (67:44)

You were. You were guttered that in New York City, that was. That was the plan.

A (67:47)

That was the idea. And the reason for that still holds, which is I'm on the west coast, so when I get up at 8am or 7am it's already midday for you guys. So I'm going to have to get up at 3am Like a baker to write. To write news.

C (68:05)

Better make the donuts.

A (68:07)

I gotta make the donuts so that I can record at 4am and get it out at 7am for you. East coasters. So I might as well move to New York to do this.

C (68:16)

Well, good. Come. Come to the universe.

A (68:18)

You got an extra room there, Jeff and the.

C (68:20)

Yeah, we do.

A (68:21)

Can I sleep in the library there? I like the books.

E (68:23)

You can come over here. It's only going to be 6:00pm.

A (68:27)

You know, that's the solution. Yeah. I do it in the night in the Philippines and it comes out in the morning.

C (68:34)

Leo's going to bug with you. But you know, just what you wanted.

A (68:38)

I'm a little nervous. The idea of writing and it'll be short, a five or ten minute daily show is kind of daunting.

C (68:45)

We'll just have AI write it for you.

A (68:47)

No, this is. So this is going to be the.

C (68:50)

You love AI. What's wrong with that?

A (68:52)

No, because the other reason I love AI is it makes us more valuable.

C (68:57)

I agree.

A (68:58)

Right?

C (68:59)

Yes.

A (68:59)

Right.

C (69:00)

So this is going to be humanity.

A (69:02)

Is the differentiator handcrafted? Darren says I should move with him to SID in Sydney. And in Sydney it's already yesterday or tomorrow. I don't know what it is.

B (69:10)

It's not now, it's not today. Whatever it is. Ladies and gentlemen, for the 2737th show, I will be 66. I would still be younger than you guys.

A (69:27)

That's depressing.

B (69:28)

Sorry.

C (69:31)

You're enjoying that a little too much, Paris.

B (69:33)

No, no, I'm not.

A (69:34)

She's laughing. But you will then have more sympathy for us today. By then we'll be so old.

C (69:41)

It'll take a while. But you'll get there.

A (69:43)

Yeah, you'll get there eventually.

B (69:45)

Only for hard life lessons.

A (69:47)

Lisa's a little younger than me and she keeps saying things like, oh, now I know why you were so tired.

B (69:54)

Oh, gosh.

A (69:56)

Now I know why I couldn't get you to go out dancing. Yeah, now that you're my age. When I. Anyway, some years ago. Our show today, brought to you by Zscaler. You know Zscaler, it is actually very appropriate for the conversation we were just having. ZSCALER understands that AI is a blessing and a curse to business. It's good news and bad news. With AI in the hands of bad guys, of course, it means more vicious and aggressive attacks, more effective attacks, more efficient attacks. But in the hands of the good guys, in the hands of your business, it means better business innovation and efficiency. Let me talk about the bad guys first. Phishing attacks over encrypted channels have increased by 34.1%. And that's primarily fueled by the use of generative AI tools. These phishing emails we used to remember, we used to be able to say, oh, if it's an ungrammatical or, you know, this is obviously efficient. Nowadays they're perfect because they're using AI to generate them. So this is a real problem. On the other hand, organizations in all industries, from small to large are leveraging AI to increase employee productivity with public AI for engineers with coding assistance, it's kind of amazing what they can do. Marketers are using it for writing. Finance is using AI to create spreadsheet formulas and much more effective spreadsheets. Companies are automating workflows for operational efficiency across individuals and teams. They're embedding AI into applications and services that are both customer facing and partner facing. Ultimately, AI is helping many businesses move faster in the market and gain competitive advantage. But there is an issue, of course, with the public and private AI that you're using in your company. We've got to really seriously think about how you protect the private and public use of AI. You don't want to accidentally exfiltrate proprietary customer information. You want to make sure you use AI responsibly, intelligently, privately. And of course, companies have to think about how they defend against these AI powered attacks. Ask Stephen Harrison. He's the CISO right on the front line at MGM Resorts International. And they use Zscalers. So that's, that's pretty good news. This is the quote. We hit zero trust segmentation across our workforce in record time and the day to day maintenance of the solution with data loss protection, with insights into our applications. These were really quick and easy wins from our perspective. MGM International loves Zscaler. You know who wishes they were using Zscaler? I'm going to ad lib a little bit from the copy because it's not in here, but it's ripped from the headlines. Jaguar announced that they still can't get the production line back on. Three weeks later, they'd hoped to reopen the day. You can't. And you know why? We were talking about it on Sunday. You know why? Because they didn't have zero trust. They didn't have segmentation, they had a flat network and once the bad guys got in, they could go anywhere. And that is a nightmare for any business. Problem is, we've been depending on perimeter defenses, you know, traditional firewalls, things like that. And, and the hope is, well, if you know they're good enough, nobody will get in. We don't have to worry. But they do get in. For one thing, you've got to create VPNs, right, so that your employees can get in that means now suddenly you have public facing IPs. That's just you hanging out on the Internet, exposing your attack. Surface and hackers with AI, they are good, you don't stand a chance. That's why you need a modern approach. That's Zscaler's comprehensive Zero trust. This is so important. Zero trust architecture in AI, Zero trust means you're segmented out. Even if your perimeter defenses are penetrated, bad guys can't do anything. They can't move laterally in your network, they can't use your apps against you. Plus zero trust protects you against zero days unknown attacks, supply chain attacks. So you get that. And with Zscalers AI, you ensure safe public AI productivity. It protects the integrity of private AI. So you're stopping AI powered attacks. You're protecting your privacy, protecting your security. This is the way to thrive in the AI era With Zscaler, Zero Trust plus AI to stay ahead of the competition and remain resilient even as threats and risks evolve. Learn more@zscaler.com security that zscaler.com security we thank Zscaler so much for supporting intelligent machines. Interesting New York Times piece. You know, we last week and I'm sorry you guys couldn't be here, but it was a fascinating thing to watch this meta event. Oh, you're muted there, Jeff.

C (75:02)

I was apologizing for deserting.

A (75:03)

Oh, no, no, no. You were on CNN for an hour. I could. I was.

B (75:06)

So I have no excuse. I went to bed.

A (75:09)

Yeah, you ate. That's all right.

B (75:10)

I ate dinner and went to bed. I'm sorry.

A (75:13)

Did you once, you know the massive demo throughout the entire meta thing, did you kind of feel bad?

B (75:19)

Yeah, I felt somewhat bad that I wasn't there, but I also don't think I would have been particularly useful.

A (75:24)

It was good. Anthony joined me. We had a lot of fun. You know, it started off really cool with Mark Zuckerberg sitting in the green room reading his script. And then he gets up, he puts on these glasses, which we don't know yet. They haven't announced anything. And suddenly you're seeing what he's seeing. You're seeing a screen in the lower right. You're seeing as he goes through the door and there suddenly he's on the stage and it's live. He's doing this live. That was the last time they worked. My God, literally the last time they worked. So the first demo they go to is with the updated Ray Bans. They did three announcements. They had updated Oakley's updated Ray Bans. And these new Ray Ban display. Meta Ray Ban display glasses. So. So the first thing he did, he had a TikTok chef on. Not my son, but another TikTok chef on. And he was going to make some sort of soy sauce marinade. And he's got the glasses on and he says, hey, Meta. Which is what you do to wake up the glasses. And she says, hi. And he says, I'm going to make this. Look at the ingredients I have. And she says, oh, looks like you got soy sauce. Yeah, that was one of like eight ingredients. Soy. Yeah, yeah, I got soy sauce. Help me make this marinade. Here's you know, what recipe. And she said, now that you've mixed all the ingredients together. And he says, no, no, no, no, wait, no, back up.

C (76:47)

He's back up. Control Z. Control Z glasses.

A (76:50)

Control Z three times. Finally, like Mark says, oh, well, never mind. And he blames the WI fi, which is not true. It's what they always. It's what Apple used to do. Oh, it's the WI fi. It turns out later Meta said, well, the problem was that these glasses were attached to development small development server. And as soon as he said, hey, Meta, it was broadcast throughout the auditorium. And all the geeks in the auditorium who had meta glasses, their glasses woke up and they all hit the same dev server at the same time. I don't know if that's true, but they said, we ddosed ourselves.

C (77:28)

Okay, that's pretty funny.

B (77:30)

That is pretty great. I mean, of all the ways for something to fail. That is spectacular.

A (77:37)

It was fairly spectacular then. And this is the picture that the New York Times published from the audience's perspective of. This is the teleprompter that Mark's looking at. It says, live demo dash dash. Good luck.

B (77:52)

Oh, no.

A (77:53)

Oh, no. So somebody's calling him. And it turns out the glasses for some reason are going to sleep each time the call comes in. So he can't pick up. So they call again. They can't pick up, so they call again.

B (78:05)

Oh, no.

A (78:05)

Four or five times they try to pick up. Never does. So the guy walks out on the stage and they try one more time. Still doesn't work. So there's another fail. I can't remember all the fails. Literally, they never worked as they were supposed to. Well, there was one kind of cool thing where Mark put on this. They call it a neural interface, but it's really just got sensors on it, this bracelet that lets you move your fingers and it senses the muscle movement.

C (78:38)

It's kind of like that Thing that Paris found around your face, right. It senses enough to be able to decide.

A (78:44)

Yeah. And you know, if you train it right, I guess it's okay. This is the New York Times reporter Eli Tan trying out the display. Mark was pretty impressive. He was able, looked like able to type, although it had autocomplete so he would kind of write. It looked like almost writing a few words. And then it would guess and he'd say, yeah, that's good. So he was able to text. The reporter said it occasionally failed to register finger swipes which is pretty much what the demo looked like. It was really rocky going the whole time. And then, and then Diplo shows up and you know, the DJ and that's I guess to give it some star power and that. I don't think that demo worked. They were trying to like play music. I guess he got the music playing so maybe I'd give him a check Mark for that. He said play some music and it played some music. How impressive. They had a completely meaningless interview with James Cameron to plug the new Avatar movie that didn't have anything to do with anything. And then finally Mark and I could tell they had planned this. This was gonna be really great. They were gonna end it and Mark's gonna go for a run. And some jock comes out all ready to run and Mark puts on his glasses. That jock puts on his glasses and we follow them out the door of the auditorium. Down in there's a bunch of other meta types wearing their glasses standing there expectantly. We're going to run with Mark. And you could tell that the whole point of this was you were going to get to see the run from Mark's point of view. They never got it working. So what we're seeing basically is we're, for some reason we're watching Mark. They had a golf cart with a camera. So he's just jogging after the. It never. It was. And I could tell Mark was like, just like he wasn't. He was as, he was as good natured as could be. But you could tell this was like.

C (80:52)

Jensen Wong would have blown up.

B (80:54)

There has to have been a huge blow up behind the scenes.

A (80:57)

I mean, God, somebody got fired. The next day I saw a report. I haven't been able to find it again, but I saw a report that the person responsible for the ability to make phone calls got fired. So you know, I'm still interested in the idea. The screen that they display is very bright. It's actually brighter than the iPhone screen. I think they said 6,000 nits the iPhone screen is 3,000 nits. They say it's cool because no one can see it. So they already. They know that there's a privacy issue. No one can tell you're looking at a screen because there's only 2% leakage. So even somebody standing across screen doesn't see the screen, but they do see the big cameras on your temples that are lit up bright red as you record. Now, the one advantage of the screen, I think, is when you're taking pictures with your new meta Ray Ban displays, you can actually for the first time see what the camera is showing. So it helps you get a better picture or movie. So that's useful. There's a limitation because meta doesn't have a phone, as Steve Steven Levy was talking about earlier. They're kind of, if this were Apple, they'd have an iPhone. ATYP, they've had an ecosystem. Meta's got WhatsApp and Facebook Messenger. So it's unclear.

C (82:21)

That's what killed it for me. I might be interested in this, but then that's all I can get to is WhatsApp and Messenger.

A (82:26)

Yeah, And Instagram, they had an Instagram icon. So great. You can look at Instagram reels while you're talking to people.

B (82:35)

What a nightmare. It's like those subway surfer memes, but in real life.

A (82:41)

Yeah.

E (82:43)

I mean, I think he was hoping for an app development platform like the iPhone. One that didn't have any apps yet either.

A (82:49)

Yeah, well, that's the argument. A lot of people said, well, you know, this is the first one and it's early and there was some cool tech now. I was impressed. It's 800 bucks, which is a lot less than the Vision Pro. You can get demos right now at luxottica stores in many, you know, the. What do they call eyeglasses, et cetera or whatever. Luxottica owns all of the eyewear stores, so some of those limited. If you go to the website, you can sign up for a demo, which you probably should do before you buy. They will go on sale on the 30th.

C (83:21)

Take it from somebody who has Google Glass still.

A (83:24)

Yeah, I think there's the same issue wearing these around that people are more and more aware of the fact that you have cameras. Did you see this story? Did we talk about this last. Last week about the woman who was getting bikini wax and looked up and saw the waxer? Did we talk about this?

B (83:40)

Oh, right, Only briefly.

A (83:42)

They had wearing Ray Ban meta, by the way, the company, her employer said no, they weren't On. Well, that's.

B (83:49)

I bet they did say that. Yeah.

A (83:53)

New York Times concludes all of this privacy stuff may not matter to Meta because it can keep afford to keep trying until it gets it right. Thanks to its online advertising business, the company made $18.3 billion in profit last quarter, even though it spent billions on Zuckerberg's other big passion that is not making much money yet, Artificial intelligence. Zuckerberg said on a recent podcast of his AI investments. If we end up misspending a couple of hundred billion dollars, I think that's going to be very unfortunate. But I think the risk, at least for a company like Meta is probably in not being aggressive enough.

C (84:31)

See earlier discussion about scale.

A (84:33)

Earlier discussion. Anyway, I don't think I'll buy these, not because of the poor demo, but I do think that they're. It's early days.

C (84:45)

What. It's what you can get if you can't test out enough of the uses.

A (84:50)

Right.

C (84:50)

You can't get directions and maps and stuff like that.

A (84:53)

I'm still chastened from wearing that little AI thing around all the time.

B (84:58)

I realize you should be.

A (84:59)

Yeah. People really don't like the idea that you might be recording at any moment.

C (85:05)

People being rife.

A (85:07)

People being.

B (85:07)

Have you forgotten the parable of the glass hole?

A (85:12)

Yeah. These aren't as obvious as the glass hole. They do light up red when recording.

B (85:18)

I mean, they should be. At least one benefit of the glass hole was that you knew that someone was recording you because they looked like a glass hole.

A (85:28)

It's also. Meta is a little bit of a problem. Right. We don't, if it were came from Apple, maybe we would trust it more. But I don't, I don't really trust Meta.

C (85:37)

Well, the funny thing about some of these, these AI link ins that are coming out now is they want you to tie your Google stuff to it so it can give you added value to it. But I'm nervous about all that.

A (85:49)

You know what Google thinks added value is? Ads.

C (85:53)

Well, no, no, I agree.

A (85:54)

Eric Schmidt saying you're walking down the street and the, and the, and you walk by it.

C (85:58)

Right, Right. From their perspective, from my perspective, added value is I can, I can look at your email and I can do things for you and I can, you know, I know what you're doing today, but I can get that out of competitive AIs only if I link them to my Google apps and the privacy and I don't want to do that.

A (86:14)

It's, it's.

C (86:14)

By the way, Google already knows it.

A (86:15)

It's. Just not just Google companies that do that. The B computer would only work with Gmail and Google Calendar. I see I get stuff all the time and since I don't use Gmail or Google Calendar or Google Contacts, there's a lot of tools I'm interested in that I can't use because they don't work with my caldev. And by the way, the fast mail which I use for my calendar and contacts uses a standard caldav and card dev, which is a standard. They could easily support it, but Google's so dominant they see no benefit addressing the Meta trust issue Here's a story from the Guardian Parents outraged as Meta uses photos of schoolgirls in ads targeting men. Instagram pictures of girls as young as 13 were posted to promote the threads site.

C (87:06)

I don't get how they were using that to promote in what way?

A (87:12)

The children's images were used by Meta after their parents had posted them on Instagram. So they were taking your Instagram posts to mark their return to school. The parents were unaware that Meta's Instagram settings permitted it to do this. One mother even said my account is set to private. But the posts were automatically cross posting to threads publicly and she's upset. The posts of their children were highlighted to a stranger as suggested threads. The recipient told the Guardian the post felt deliberately provocative and exploitative of the children and the families involved. Father of a 13 year old who appeared in one of the posts said it was absolutely outrageous. The images were all of the schoolgirls in short skirts with bare legs or stockings. Meta, not what you want, knew what it was doing and you know this is a cautionary tale. Stop posting to Instagram if you want it to be private, even if you have a private account. Meta said that's just a recommendation tool. The images shared do not violate our policies and our back to school photos posted publicly by parents.

C (88:29)

That's a good way to ruin any trust you might have had with parents.

A (88:32)

Exactly. Yeah, exactly. A 37 year old Instagram user from London received the posts and asked to remain anonymous. Said over several days I was repeatedly served meta adverts for threads that exclusively featured parents images of their daughters in school uniforms revealing their names. Thank you Meta.

C (88:56)

In this time of Epstein, can you read the room?

A (89:01)

I, I, I really, I never post Instagram anymore or I mean I just, I feel like these, these sites are not your friend. Obviously in your generation, Paris, you probably, you don't have a choice, right?

B (89:14)

You have to use Instagram, always be posting. I mean I know some People who don't use Instagram, but they're few. Kind of the way that I keep friends. I used to own. I used to be maniacal, but only having a private Instagram account. But I changed it when I was getting more public because I kind of got overwhelmed with a bunch of people I didn't know trying to add me. And I was like, well, I guess I'll just make it like a quasi.

A (89:40)

Public and post with, you know, with that in mind.

B (89:43)

With that in mind. Yeah.

A (89:44)

Yeah. No, I love your posts. The. The. The posts you did of the soapbox derby. Those are great. Those are fun.

B (89:52)

But yeah, it's. I always post the idea that the world is watching.

A (89:59)

Well, parents of young girls, maybe you should do the same. It's sad because people really do want to share. Mark's not wrong that people want to connect. It's just that's. That's the wrong place to do it, I think. I don't know. There is nothing.

E (90:15)

There's a lot of ego getting in the way here because he can just. You do Facebook and still be a billionaire.

A (90:20)

You know what I mean? That's right. That's why he says, I don't mind If I lose $100 billion on this moonshot. I get plenty more where that's coming from. Yeah.

E (90:29)

So it's ego at this point. He just wants to be the next chance.

A (90:32)

And that's why we're talking with Stephen. That's why there is a concentration of this in the oligarchs, the people with lots of money. No one else can do it.

C (90:41)

When I was on the board of Nick Denton's company before Gawker and the crash in 2000 came, he had raised a bunch of money. I said, nick boy, best thing he could have done is raise a lot of money. He said, no, Jeff, you're completely wrong, as he tended to do. And he said it, let the CEO play with everything. And one of the smart board members with a wonderful name, Ricky Tata. I love that name.

A (91:03)

Sounds like a Rudyard Kipling poem. You're a better man than I am, Ricky Tata.

C (91:09)

So Ricky insisted that a separate bank account be set up that you had to basically ask to get money out of it to do something good. And I think some scarcity of this goes back to the question, when you have all the money in the world, literally, you can spend it. When I worked at Time Income and I got an argument what system we were buying because I was doing boring things like that. And they said, basically, if we're. If we're timing and we're spending the most money on it. It must be the right thing to buy, right?

B (91:39)

Oh, no.

C (91:40)

And there was no discipline. Yeah, right, exactly. So I think Zuckerberg has too much money.

A (91:50)

Have you tried Hux?

C (91:53)

I, I, this is what this is. We're going to have them on next week, right?

A (91:58)

Oh, are we? Is that set?

C (92:01)

I saw it in the schedule.

A (92:02)

Oh, good. I should probably look at the schedule once in a while.

C (92:07)

So these are three people who were part of the original development team for NotebookLM whom I met when I, when I, Stephen Johnson, showed me Notebook Lm really smart. Obviously redundant to say that they were at Google, but they left and I worried at the time. Oh my God, what's that? I love Notebook lm. What's going to happen to it? Well, they've been developing tons of stuff at Notebook Lm doing a great job. Job. But at the same time, they have just released their new company. Is it Hux or Hukes?

A (92:31)

Well, I'm thinking it's, it's short for Huxley. Why would I think that that they're naming it after Aldous Huxley.

C (92:39)

But why would.

A (92:39)

Who wrote the Doors of Perception anyway? I don't know. What else would Hux be? Luke's. I don't know.

C (92:48)

I don't know. Hux AI.

A (92:51)

So this, it's an app, right?

C (92:52)

Right.

A (92:53)

And basically is like Notebook lm. It gives you a podcast. Multi voices, by the way, not just two. Reading the news.

C (93:04)

You want, you want to hear mine?

A (93:06)

Yeah.

C (93:10)

Come on, come on. Oh, here we go.

A (93:14)

By the way, this music is really irritating. And it's the same on all of these. Yeah. Hey, Jeff, Good evening. It's a cloudy one out in Bernards. Hopefully you're staying comfortable as the day winds down.

B (93:25)

Glad to see you popping in just after six.

A (93:26)

Could you slow it down? I'm sorry, I can't.

B (93:30)

What level of speed is that? Oh, boy.

A (93:34)

Two. Now we know how you read all these books.

C (93:37)

That's exactly how I do it.

A (93:38)

Yeah, now I understand. I can't listen. Do you listen at high speed, Paris?

B (93:44)

No, I don't.

A (93:45)

Thanks for the feedback. We always appreciate hearing from you.

C (93:49)

It was listening.

A (93:50)

I'm glad to know our approach is work out for you. Let's get back to the news. All right, moving along. And a few fresh twists in the art world. Perfect mix, Right?

B (94:06)

Let's head straight to today's news updates and see what's next on the horizon.

C (94:11)

Let's dive into the day's most intriguing.

A (94:13)

This is Why I want to do a daily news show written and delivered by a human being.

B (94:19)

Yeah, I think that that's neat.

A (94:20)

Here's what else is making headlines before we wrap up this segment. It's ironic because when I saw this, my face fell because I thought, oh, this is exactly what I was thinking.

B (94:27)

Data centers in the US with Oracle and SoftBank as part of the massive 500 billion dollar AI infrastructure push, they're aiming for nearly 7 gigawatts of capacity and over $400 billion invested. This all broke today, right? This all broke today.

A (94:45)

Are expected to create 25,000 jobs and lock in American leadership in AI. That's a press release. So this is the, this is. The difference is we said, well, wait a minute, hold on, you know, let's.

C (94:57)

Go into the, the details here.

A (94:59)

Yeah, yeah. And I think that that's, that's what I would try to do with the Daily show as well as I would only do a few stories, maybe two or three, but try to give you some perspective, not sentences. Me saying, why does it have to be the morning?

C (95:11)

Why can't it be the evening?

A (95:13)

Well, I just think that's the way people listen. They listen in the morning as they're getting ready, they're, they're preparing themselves for the day. The evening, I guess. I don't know. You like evening news? I guess we watch tv.

C (95:24)

Well, just for your sake.

A (95:26)

Just for my sake.

E (95:28)

The problem is that evening news, like most people would probably listen to that the next day and it might be old news already by that time.

A (95:34)

I don't want to be old. You know, Kagi has created a couple of new apps and I was pretty excited at first because one of them is a news app. And I thought, oh, you know, this is going to be great because I've already taken it off the front page of my. Because they say, you know, it's going to be apolitical and you get to choose the subjects and so forth. My problem is it doesn't update but once every like two or three days.

B (96:02)

So actually it's taking a long time to read all that now.

A (96:05)

Yeah, I don't. Anyway, actually this is updated now, so. But this morning, it was two days ago, they were saying like, you know, stuff that was just out of date. And I didn't think that. Yeah, okay. Trump sets $100,000 H1B application fee.

C (96:19)

That's old news.

A (96:20)

That's old news. So I like the idea and, you know, maybe I'm gonna give them the feedback and maybe they will get it Updating more, more quickly. It's pretty cool. You pick the subjects you want. I have. Here's the AI section. Now they don't read it to you. There's the Stargate story, which we did. Websites offer three AI creation tools. I don't know what that is. Alibaba, we talked about this QN3 model.

C (96:47)

So when you go to a story, I'm now at the story about the Secret Service.

A (96:49)

It's an AI summary, right?

C (96:51)

300 thing. Yeah. That is a summary.

A (96:53)

It's like perplexity where it gives you the original sources, gives you a summary. I think this is an interesting product, kind of hucks without the voices. So I guess Hux would be good if you, you know, you're driving and you, you know, you wanted it to be read to you.

C (97:07)

What Hux also promises is I didn't give it access to my calendar and my email. It would say Leo sent you another email reminding you to do what you promised to do, but you haven't done.

A (97:21)

This is another one of it only works Gmail and Google Calendar. And that's why. Another problem I had. Although maybe you shouldn't tie it to your email. Email.

C (97:30)

Yeah, I'm a little iffy about that.

A (97:33)

Yeah. And also email would be privacy reasons.

E (97:35)

It would also be inadequate. Right? Like because it would need to know your text chains. Also because if you move conversations between mediums, they won't know any about that.

A (97:44)

I'm telling you, ecosystem is really important. And I think this is my reaction to the meta glasses as well. And this is why I think Apple has a leg up for people who are in the Apple ecosystem.

C (97:56)

And if you trust me and Google, same thing.

A (98:00)

If you're in Gmail, you know, then.

C (98:03)

If you're an Android phone, you're in Gmail, you're in Drive.

A (98:08)

Which is why I don't use an Android phone. Because I don't want to be a living la vida Google so much. I trust Apple.

C (98:16)

The water is fine, Leo. The water is fine.

A (98:19)

It's not as smoothly integrated. Apple's really focused on that. But admittedly I'm a sheeple. I'm now stuck in the Apple ecosystem. And let's hope Apple's dedication to privacy is genuine.

B (98:33)

I will say, Jeff, you. Our group chat for this podcast is the first group chat with an Android phone I've been in that isn't completely busted. I'm in another group chat with an Android phone and it is so broken. And I assume it's broken because it is related to Android, but I. Because I haven't had this experience before and I had assumed it was fixed due to rcs, but that's my main.

A (98:59)

I have had no trouble with rcs. In fact, I haven't had any problems since Apple put RCS on the. In the iPhone. It's been great. My daughter uses an Android phone. It's indistinguishable from Apple.

B (99:10)

Wait a minute. Is RCS only a feature in the latest iOS Apple? Is that correct?

A (99:16)

Are you way back on?

B (99:18)

I'm not, but some of my friends. Some of my friends are the people who aren't on Instagram and they have, like an old iPhone that's still like an iPhone. Little touch button.

A (99:29)

Yeah.

B (99:29)

And so that's probably it.

A (99:30)

There is a setting you could go check in their essay, in their text message, in their messages settings.

B (99:35)

Yeah. But I don't. I don't think they.

A (99:37)

If they don't see it, they don't.

B (99:39)

They don't have the updated iOS that.

A (99:41)

Might be the problem. If. If everybody's up to date, it works pretty well, actually. We, the. The three of us, plus Benito and Anthony are in a group text. This works quite well.

C (99:50)

That's right.

A (99:51)

It's very busy.

C (99:52)

Benito even harassed Leo with the Craig theme.

E (99:56)

It even worked better.

B (99:58)

That made me cackle. It was. They mentioned. Someone mentioned Craig in our group chat, and within seconds, Bonito sent the Craig, Craig, Craig Newmark theme to the group chat, which I guess Resol bolted and it automatically playing on Anthony's phone.

A (100:17)

Did you send that from an Android phone or an iOS phone?

B (100:21)

Did you just keep that thing on you?

E (100:22)

I sent it from Mac, actually. I was on my computer, so I just opened up messages and I dropped.

A (100:26)

The file in because it's an MP3. Yeah. That's interesting.

B (100:29)

It sounds beautiful in this version because you can hear all the human voices.

A (100:33)

Yes. Much better than usual.

E (100:35)

Did it autoplay for anybody else or just Anthony?

A (100:37)

Did not auto play.

B (100:38)

It did not auto play for me.

C (100:40)

But it did for Anthony.

B (100:41)

I think it's very funny that Anthony had headphones in, was in like a.

A (100:45)

Different room or something, and automatically heard Craig. That's why that is scary.

B (100:52)

He's all around.

A (100:52)

I want you to know, because of that, I showed real restraint. I was really ready to put some other sound effects in there. Then I thought, no, I'm not going to torture Anthony.

C (101:02)

I can share with you the MP3 of my call into Howard Stern this morning.

A (101:05)

Morning. Why don't you hold that for a moment? And that's next on Intelligent Machines. Jeff Jarvis, Paris Martineau. Benito Gonzalez, our producer and technical director, also chiming in. It's great to have all of you on the show and all of you watching, especially our club members. Thank you. It's becoming more and more important. We made a big announcement today, which I will mention, as you may or may not know. You know, we've been doing our own ad sales for quite some time, I think. I think when we first started doing ads on Twit 15 years ago, we used a company called PodTrack. They were great. But ultimately we took all of our ad sales in house. And Lisa, in the last couple of years, has been doing it all on her own. We've had. She's had a little. I shouldn't say that because she's had a little help, but she's pretty much the main salesperson. We have decided to give Lisa a break. Is this appropriate to say, Lisa, we have decided to give you a break. And we have partnered with a company called Airwave. We have a strategic partnership. I can say this right. It's public now and they're going to do our ad sales going forward. So we're very excited about this because it means Lisa will not be working seven jobs anymore. She'll only be working six. So that's a big change for us. However, it's a little bit of a risk because no one sells TWIT quite as well as Lisa. She's been really great with the ad sales. So we know we're taking a little bit of a chance. So that's why it's so important that you show your support for TWIT right now by joining Club twit. Yes, we are. We do have ads, but if you pay the 10 bucks a month to join the club, you won't hear them. You get ad free versions of all the shows. You get access to the discord. You get all the special programming we do just for club members. I think it's a really good deal and most importantly, it really supports what we do. So if you like the shows and you want to keep hearing them and you want them to expand, if you want that daily News show, this is how you vote. Twit TV Club Twit. Two weeks free trial. If you want their family members memberships, corporate memberships, but it really makes a big difference to us. Twitter, TV Club Twit and.

C (103:22)

And sponsors who want to still sponsor this show, where should they go?

A (103:28)

What? Yeah, what should people do now? Should they still. They should still go to advertise@twit.tv. where should they go at least if they want to sponsor the show. She stopped paying attention. I don't know. I don't know what's going.

C (103:43)

She's already taken some time off as well.

A (103:45)

She's taken. She's on the beach.

B (103:47)

Us or any. Does this mean anything for us?

A (103:50)

No, not at all. It'll look exactly the same.

C (103:53)

You're still going to do.

B (103:54)

Will you still be doing in live ad reads or will they be okay? So I will have a break to go eat a spoonful of peanut butter.

C (104:03)

How does this affect carrots?

B (104:06)

That's what I'm always asking with everything is how does this affect me?

A (104:12)

There will be. We'll continue to do what we have been doing when we don't sell it. Ads, which is those ad avails sometimes go to direct ad, what they call direct ad insertion, which is a third party ad that gets inserted on the fly as you download so you don't.

C (104:26)

Hear Leo's delightful voice.

A (104:28)

Some of them will be my voice, some of them won't be. Yeah, and then we still hope. I'll tell you what, if you don't see me doing the ads, join the club.

B (104:38)

Join the club so that I can have a single scoop of peanut butter while Leo reads the ads.

A (104:44)

AdvertiseWit TV Lisa has confirmed that very important. Oh, I know why there's a delay. She's listening. She's not listening in the room, she's listening to the stream. So she just now heard us.

B (104:56)

Are we on an ABC Kimmel esque delay so someone can come in and cut the stream if we see something here?

A (105:02)

You're our Guillermo. That's right. Advertise at TWiT TV. Same as always. We love having you approach and say I'd like to. That's what. Actually it's so funny. This is what happened to our sponsor for this segment on the show Pantheon. They said, we'd like to buy ads. And we said, well, as a matter of fact, did you know we use you? I guess the people buying the ads had no idea. But Pantheon has been vital to our workflow for some time now. You know, it's vital to your workflow too. Your website is your number one revenue channel. But you know, when your website's slow, people don't stick around. Or if it's down or it's stuck in a bottleneck, your website could be your number one liability. That's why we're so happy to be with Pantheon. Pantheon keeps your site fast, secure and always on. That means better SEO. More conversions, no lost sales from downtime. It's not just a business win, it's a developer win, too. Ask our web engineer, Patrick Delahanty. He loves Pantheon. Your team gets automated workflows, isolated test environments, and zero downtime deployments. So for Patrick, that's no more late night fire drills. No more, well, it works on my machine. Headaches. Just pure innovation. We love Pantheon so much. It's not just our website. It's our entire backend, our entire workflow. Running on headless Drupal, running on Pantheon servers. That's why it's more important than just a website being down. We wouldn't be able to publish. We wouldn't be able to do anything we do without Pantheon. And they are so reliable. You'll love Pantheon. Marketing can launch a landing page without waiting for a release cycle. It's just modern developers can push features with total confidence. Your customers, they just see a site that works 24. 7. Pantheon powers Drupal and WordPress. Sites that reach over a billion unique monthly visitors. Visit Pantheon IO and make your website your unfair advantage. Pantheon, where the web just works. He says the number. Patrick's on the. In the discord. He said, the number of times our site on Pantheon has gone down and I've had to fix it in the middle of the night. 0. 0. Yeah, we love Pantheon. So it's kind of ironic. They said, hey, can we buy some advertising? We said, yes, gladly. We love pantheon. Pantheon.com IO you were on Howard Stern.

C (107:28)

When, this morning?

A (107:31)

Yeah. You know, I could play it from my computer here.

B (107:33)

Howard Stern Weekly lately.

C (107:36)

I was on last week. Yeah.

A (107:37)

Does he know you? Yeah. Is it like. Are you, like, some stranger calling in.

C (107:41)

No, no.

B (107:43)

He's been in the studio.

C (107:45)

Yes. I talked about my prostate in the studio.

E (107:48)

I'd recommend again.

B (107:49)

We talked about his public park and Howard's private parts, famously.

A (107:55)

All right, let's. I guess I have to log into SoundCloud.

E (107:58)

I don't know how much of this we should be playing.

A (108:00)

Is it Filzit? Oh, because you think Howard show.

E (108:05)

I think they're probably going to come after that, Right?

C (108:07)

Oh, okay.

B (108:09)

Give us a summary.

C (108:11)

It was calling by Kimmel, and Kimmel was brilliant. Last night, I think Howard.

A (108:16)

I was very proud of Howard. Howard said, I am canceling my Disney subscription, which, by the way, carries a huge amount of weight. I imagine millions of his listeners did the same.

C (108:29)

His lead.

A (108:29)

Yeah. And is he gonna renew? Now here's the question. Are you gonna renew your subscription now.

C (108:35)

That they brought Kimmel Back. He said he would. Yeah.

A (108:38)

Reward them in the same way you punished them.

C (108:40)

Well, I even said, when I was on, I said that unlike Mel Karmazin, who stood by Howard back in his fcc.

A (108:48)

Oh, man. Howard had plenty of trustees.

C (108:49)

Right. That's what's forgotten. Here is the fcc. He's been bad before. I wrote a Nation cover story called F by the FCC about Howard. He was very appreciative of what I had written in the day. It was really touching that he felt like he was so alone and that I would write things that were in support.

A (109:07)

And his book Private Parts was really good, describing his troubles because he was a shock jock in the year, you know, when shock jocks just began, and he did some pretty offensive things. But that's a perfect test of free speech. The difference is the FCC does, you know, very famously control the public airwaves, and you can't say the seven dirty words, George Carlin's words. He's on satellite now, which means he can do anything he wants.

C (109:36)

That's why. And so I argue that's a slice that shouldn't have been taken out of the First Amendment in the first place. And it was because of the early days of radio. There's a moral panic about the technology of radio invading the home.

B (109:52)

Wow. I've forgotten. I've forgotten.

A (109:54)

Terrifying. My robot face just emerged. So I'm not sure I disagree. As somebody who worked in radio for 50 years and, you know, we don't do expletives on this show. We bleep them out as if this was broadcast.

D (110:10)

I know it's hard, and I kind of.

A (110:15)

I think that's a good thing, especially in this day and age where there are plenty of places where you can get an expurgated speech like podcasts. Right.

C (110:25)

But that's also the definition of offense. One of my favorite bits of reporting way back in the day was that the FCC put its largest fine up to date then in its history against awful use of whipped cream on Foxes Married to Americans sitcom which didn't last very long, and that there were countless. That the world was in an uproar. Countless complaints. So I did a FOIA and I said, send me all the complaints. They did. And all but two were identical. Written by the same place. You just. People just hit a button, button, button, button. Right. And so only three Americans had bothered to take the time to write that they were upset about the show. And on that basis, the FCC did this huge fine. And now the FCC is going after things. So. So Howard is trying to avoid politics. Now, it's very. I heard him, just like you, Leo. He said, I'm out of politics now. My readers don't want it. My listeners don't want it. I'm not doing it.

A (111:24)

And in fact, to some degree, I think people want a respite. It's not that they're apolitical, it's just that they. It's everywhere. And they would just like to listen to us talk about AI.

C (111:32)

But in this case, he had. He called Jimmy. He had to stand by Jimmy. They're very good friends. He told Jimmy, you know, I know. I've been there. I know where you were. And I asked Howard whether he ever thought of giving up. And because it was very frustrating for him, and he said no, never, never. He would keep fighting.

A (111:51)

He was pretty offensive. I don't. I thought. I actually did not like his material. I thought it was quite offensive.

B (112:01)

People can be offensive on their own.

A (112:04)

Well, here's my. My attitude. I think on the public or waves, it's appropriate to make it safe for everybody. And I think that. That, especially now, is doable because we have podcasts and the FCC does not. Let's hope they never do, regulate podcasts. Independent podcasts like ours were not owned by a big company that's trying to do a merger. No, there's no parent corporation is going to tell us what to say or what not to say. And there are literally millions of us. So if you. There's plenty of places for free speech. I think the public airwaves are something different. You know, one of the things radio stations have to do is show commitment to their local communities. They have to maintain a file of things they've done. They have to run public service announcements. That's why you hear them. They have to. I think that's all appropriate. They're using public resources, limited public resources.

C (113:02)

We're losing the. So. So now we have. The Ellisons are buying. Have bought cbs. They're going to buy cnn. And that's why I was on CNN last week, because I was the one. I was the one who could say, I know.

A (113:15)

I love your line. I loved your line. It's the dying. What is it? The dying breaths of the old media.

B (113:21)

And then Scott Jennings looks straight to camera and smirks.

A (113:25)

Well, it's funny because I made a gif of that. You're, in a sense, sitting in a studio that is part of this.

C (113:32)

Oh, I said. I said the odds are coming next for this company. Yeah, the elephant on the table, you.

A (113:39)

Know, frankly, let them Acquire all this dying media. What everybody should do is support independent Internet based media because this is an opportunity. There are play. Look, there are so many great podcasts from all sides of the political spectrum, including Charlie, Charlie Kirk's podcast. He had every right to do what he did in podcasting. I think that is a strength and I'm a, you know, I'm a massive supporter of it. Now we decide as a company not to put profanity on our podcast for the same reason. Yeah, yeah, we make that choice. We want kids to be able to listen. The parents feel comfortable with their kids listening.

C (114:20)

Do I have to listen to that show again?

A (114:22)

Yeah, parents. I know. And in fact, when somebody else on the radio, we did a twit. We did a twit with Amy Web and you know, Harper, Harper Reed. You know, Glasses, Glasses. Harper Reed. Sorry, Harper. And, and they were very passionate in what they were talking about and they started swearing up a storm. And, and we debated afterwards, should we bleep this? It. I, I thought it would be an insult to the audience to go bleep, bleep, bleep, bleep all the way through it. So what we did is we put a disclaimer at the beginning saying, you know, usually our stuff is we, we bleep out explicitly.

B (114:57)

You gave them the Ed Zitron disclaimer.

A (115:00)

Yeah, yeah, yeah. And I think from time to time that's fine. I did, you know, I did hear from a parent who said, you know, I'm listening to the car and oh yeah, yay. And my kids are hearing this and I'm not happy. But I, so that's, you know, but that's our decision and I think that's great that we have, have, we can make that decision.

C (115:18)

But here's, here's, here's the issue. So, so media go. I wrote a very depressing post this week thinking it's all lost and, and CBS and CNN are lost. So where do we go? Dissent? Well, we could have gone dissent at TikTok, but that's being bought by the Ellisons and Murdochs too. And then, well, let's make a podcast and put it on YouTube and I live la vie to Google. But now just today YouTube said, oh, never mind. We're going to put back on the election tonight.

A (115:46)

Disinformation.

C (115:47)

Yeah, and the disinformation people. And the pressure is going to be on them.

B (115:52)

These are all platforms, YouTube included. Part of the reason why we can't swear. Say things. You get demonetized if you have too much Profanity, or if you say certain words relating to sexuality or violence, it's. It's bad all the way down.

C (116:08)

Yep.

A (116:09)

Well. And again, the beauty of podcasts, there are RSS feeds. Nobody controls them. It. Nobody gets to say. No platform gets to tell us anything about our podcast. The advertisers are the biggest threat to us is one of the reasons I push so hard for the club. Because fortunately, people who advertise on our show kind of understand what they're getting, I think. But I could see pressure from advertisers saying, yeah, we don't want you to do. Sometimes they have. Have, you know, clauses in their contracts say, you know, we don't want to be next to any political content, things like that. That's why the club, to me, I mean, the. The platonic ideal of media would be an independent voice supported by its audience.

C (116:59)

Well, and. And in a way that can't be taken down. And this is where, you know, I quote you. It's the lesson of Mastodon that.

A (117:07)

Yep.

C (117:08)

Federation, Federation and protocol. Mike Basnick. Protocols, not platforms.

A (117:13)

Although I have to say, as a Mastodon admin people may not know, I run a instance for our. For you, our audience called Twit Social. You're welcome to join. All you have to do is mention that you listen to the shows because there's lots of spammers trying to get in and don't know anything about the show. So just say, you know, Leo sent me or whatever. All of us Mastodon admins are grappling with this Mississippi ruling by the Supreme Court that Mississippi can block social networks that don't age Verify, and Mastodon has no facility for me to do that. What I do in our rules now, as a result is I put you must be 18 to use twit Social. And then I have a little box that Sundays, I'm over 18. But that's just the honor system. I don't think it would satisfy Mississippi be. So we're all grabbing.

C (118:00)

Maybe you want to edit that out of the show when the trial comes.

A (118:03)

Well, no, it's the truth.

C (118:04)

You admitted on September 24, 2025.

A (118:10)

Oh, okay. Do they Advertisers, do they. They all the contracts. Okay, but do we. Should I be worried about what we say on a show?

B (118:24)

Absolutely. Absolutely. It's in every single fucking context.

A (118:27)

So everybody please join Club Twit so that we don't have to worry about that.

C (118:32)

Yeah.

A (118:33)

Join the Club Twit TV Club Twit. I could see at some point us saying, you know, we can't, you know, we can't speak truth to power or whatever because advertisers don't like it. So far we haven't had that problem though, I have to say. So Paris needs to be freed. Why do you need to be freed, Paris?

B (118:55)

Darren Oakley said, I want more swearing. Where do we sign up for that? Free said free Paris Martineau. I respect Free me. And I said, I will say. One of you guys pointed out whenever we got in a group chat, us and kind of the producers, they were surprised to find out how much I swear one of you guys was because I obviously don't do it in the show because I'm not allowed to, which.

A (119:17)

We'Re proud of you. This is good for you.

C (119:19)

We know how hard it is.

B (119:20)

I am a new news girl at heart.

A (119:23)

Does your. Do your parents say, Paris, watch your language. Did you have a swear jar as a kid? No.

B (119:29)

No, I don't think I. I think I waited until long enough that they don't care.

C (119:35)

My mother couldn't stand it. And then finally there came a point later in life where she started doing it for my benefit for the joke.

A (119:41)

Oh, that's funny.

B (119:42)

Yeah. I will say I. So I take note when my mother. My father swears occasionally. When my mother swears for emphasis based on something I'm like, oh, I've gotten into.

A (119:53)

But if you don't swear, then when you use this strong language, it carries more weight. I occasionally will use strong language.

B (120:00)

My constant use of profanity reflects the reality that we live in, which is more extreme and polarized than ever. And anyway, I'm doing a performance art piece.

C (120:11)

Yes.

A (120:12)

Is that coming up soon? Can we, can we go see?

C (120:14)

I think we have to do this.

B (120:15)

It will be part. So this is gonna be part of our 24 hour twit livestream that you guys will tune into on New Year's that I'm just going to keep talking about like it's going to happen until it actually occurs. We could have a swear out.

A (120:29)

Lisa proves. By the way, Lisa approves.

C (120:32)

She.

B (120:33)

We need to find a proper way to fund late night.

A (120:37)

We should do a late night show where anything goes.

C (120:40)

I think so.

B (120:40)

We should do one twit late night. Yeah. Where we all get a little stoned and swear. I think that could be fun.

A (120:46)

That could be the gaming thing that you wanted to do. We're all planning a game game that could be good. Anthony, set it up. Okay. I am not going to participate. I don't use drugs or alcohol and I don't swear. I am basically the Pat Boone of podcasting.

C (121:02)

You don't swear in your.

A (121:04)

No, because I've been a broadcaster for 50 years.

B (121:07)

You can't turn it off. He's always talking.

A (121:10)

You don't want to swear too much because it might leak in and, you know, it was only two years ago that I stopped doing a radio show. I did my whole life. I was on the fcc. TV or radio?

C (121:23)

Yeah.

A (121:24)

I was beholden to the fcc and I didn't feel like, oh, great complaint. No, I did tell a pretty racy joke on the screensavers. The FCC didn't complain, but management did. But that's a story for another day. Okay, pick some stories because we're running out of time. Believe it or not. I can't believe it.

B (121:55)

I just really, like. I didn't pick this, but it did make me giggle to see 84, which is fat Bear week is back. And the.

A (122:05)

All right, we'll do it now. Let's do it now. That was going to be my pick.

B (122:08)

I'm so sorry. I'm so sorry.

A (122:09)

We did this last year. Year, I think. Didn't we talk about this last year? So there is a live stream from the Katmai National Park. It started the 23rd, started yesterday. The reason they do this is. This is when the bears come out to eat the fish. And let's see. Let me see if I can. I'm getting. Is a lot of stories I want to link to the actual camera. Ah, here it is, Fat Bear Week. So you can watch these live streams and vote for your favorite bear. Here's. Let's look at some. Caught my webcams. Here's the Brooks Fall cams. We've done it. Oh, not available. All right, this is from the National Park Service. Maybe. Maybe they're. Maybe the budgeting is. Oh, man.

C (123:02)

Oh, well, so much for this.

A (123:04)

Oh, man. Maybe.

C (123:04)

Thanks for taking us here.

A (123:05)

Paris National Park. No, I was. I'm the one who bookmarked it. I really wanted to show this. Last year we showed this and you could. You would actually get to see the bears. They're feeding right now, getting ready for their hibernation. I don't want to download it. I want a live stream of bears eating salmon.

B (123:27)

Yeah, that is unfortunate.

A (123:28)

Do you think that they lost their funding? Live stream not available. River watch live stream not available.

B (123:38)

Oh, I'm just gonna search Fat Bear live cam. Now.

E (123:44)

Be careful.

A (123:45)

You are exiting. Then.

B (123:48)

I'm looking at live fat bears right now. Guys, that worked. I can't believe it.

A (123:53)

Ah. Did you go to explore.org Fat Bear.

B (123:57)

Live Cam now and the first thing that came up was Brown Bear, Brooks Falls and Cat May National Park. I just put it in the chat. I can't believe that worked.

A (124:07)

Oh, and here are some fat be you can vote for.

B (124:10)

The bears are looking right at the camera right now.

A (124:13)

It's still. I know, it's so great. So you could see. This is before and after. This is, you know, after the. After eating all the fish. So, so what did you search for?

B (124:23)

I just put in. I, I search for fat bear live cam now. But it's also in the chat.

A (124:28)

Okay. All right.

B (124:31)

There's a bear eating some salmon, being surrounded by birds. This is great radio. People who are listening are gonna love this.

C (124:40)

The bear is gonna have us take them down.

A (124:43)

Oh, yeah. You gotta explore.org okay. Oh, look at this. This is great. And it's even got sound effects.

B (124:49)

Oh, it's of course now panic. Oh, it's panning to a different bear. That's good. You can see a Baranis.

C (124:59)

Yeah, we're clean live.

B (125:01)

Yeah.

A (125:01)

Watch now. We're clean. Oh, these are the. These are all. Look at the rangers. They're lobbying for their favorite fat bear.

B (125:08)

Rangers that have jobs.

A (125:10)

Yeah. Oh, look at these fat bears. Should we just, you know, forget the show? Let's just watch.

B (125:18)

This could be 30 minutes of the 24 hour live stream. We could also. I'm thinking I have another 30 minutes where we watch the cutlery corner. Got it all mapped out.

C (125:29)

So bears don't eat birds?

A (125:31)

No, the birds eat the leftovers.

B (125:33)

Oh, I think the birds are too small to be of concern to the bear.

A (125:37)

Well, the bears want a nice big fat salmon and that's what they're hunting for. And then there'll be leftovers which the birds apparently enjoy, I guess. I don't know.

C (125:48)

I'm not dunking his head.

A (125:49)

I'm not ranger Leo or anything. That's just my guess.

C (125:52)

I think. Got one.

A (125:52)

Whoa.

C (125:53)

Yeah, got one.

A (125:54)

There's a little jumping. Oh, he did. No, he didn't. I imagine the six there. You see it? Did you see the fish jump? The success rate on this is probably fairly low, even though these rivers are jammed with salmon this time of year.

B (126:07)

Wow. Darren Oakley made a AI image of me on a fat bear, quickly wearing a backpack. And I do respect it.

C (126:16)

Oh, that's good. That's good.

B (126:18)

Yeah, it's a good one.

A (126:19)

This is good. You look like. What was that book, the Golden Compass that she wrote?

B (126:24)

I do, yeah.

A (126:25)

She wrote a bear this week in Fat Bears. I'm gonna leave the river going in the background.

C (126:32)

It's better than the construction.

A (126:34)

Yeah. Other stories. Professor Jeff Jarvis, do you have one that you want us to check out? You always seem to have something.

C (126:45)

Well, we could look at an academic paper. My boyfriend is AI. A Analysis of Human AI Companionship in Reddit's AI Community.

A (126:54)

I. I bookmarked this too.

C (126:56)

From the MIT Media Lab.

A (126:58)

Yeah.

C (127:00)

So there's clusters around dating and romance, or community support. Or meet my AI partner. They put up pictures of their AI partners. They put up pictures of their rings.

A (127:12)

Oh, dear.

C (127:14)

Upon getting married.

B (127:15)

Oh, yeah.

A (127:16)

Do you think people are actually marrying AI?

B (127:20)

Yeah, yeah.

C (127:22)

If you take.

A (127:23)

Absolutely.

E (127:23)

People married pillows and dolls and stuff.

A (127:25)

So I know they married pillows.

B (127:27)

Some woman marry a. A waterfall.

A (127:32)

You marry a waterfall.

B (127:34)

I do believe this is correct.

A (127:37)

I married a waterfall. So here's an example of user generated visual representations of AI companions illustrating the spectrum of artistic styles within the community. From left to right, a photorealistic portrait emphasizing human like features, including glowing flowing hair, an off the shoulder sweater, and a nice steaming mug of coffee. There's also a stylized embodiment with anime inspired aesthetics and a romantically staged couple photograph depicting the user AI relationship. Have you ever asked an AI to draw a picture of you?

C (128:21)

All right. Now you're going to try it, aren't you?

A (128:23)

I've done it.

C (128:24)

You have?

A (128:24)

And Reddit's loaded with them. And it's often very depressing. Often a loser in his jammies in his mom's basement. I don't know. Should I? It's gonna take a while to do it, so we should do other things while I do that. That's a good story.

B (128:41)

My boyfriend is AI I confused a waterfall with. There are two women who married rollercoasters. My apologies to you.

A (128:49)

Big difference.

C (128:51)

Yeah, big difference.

B (128:52)

I know one's got a lot more ups and downs. Downs to it. One's got just a down.

A (128:59)

My relationship. Oh, Chat. GPT says I don't actually see you. No camera feed, no secret metal Polaroid. What I've got are the stories you've shared, the vibes in your words, and the imaginative flourishes you've asked me to play with. So if I draw you, it's less a photograph and more a mythic portrait stitched together. Oh, Dennis says you want me to turn this poetic sketch. Yes. Into a actual viz visual image, so I won't read you what it suggested it was.

B (129:30)

What did you put as the prompt?

A (129:32)

A little nice Just, I said, draw a picture of me as you see me. This is what it said. A broadcaster's voice frozen mid laugh. A podcaster's with headphones like a crown of orbiting moons. Maybe a chestnut tucked into your pocket. Maybe kettlebells and clubs leaning against your chair like medieval weaponry. A glowing ipod or mike as your scepter. A swirl of Van Gogh sunrise beside you. A science fiction paperback scattered. These are all things I am not.

C (130:03)

Going to be replaced.

A (130:04)

And a cast iron skillet hovering just above your shoulder like a halo. Forever sizzling with future hash browns.

B (130:12)

Forever sibling with future hash browns is really haunting.

A (130:15)

Actually that's the picture you're never able forever.

C (130:18)

You're never getting away away from them. It's like a cloud over your head of sizzling.

A (130:22)

It is basing this on things I've asked it about.

B (130:25)

You've asked it about hypothetical, unattainable hash browns. It's not.

A (130:32)

I asked it some time ago for hash brown recipes and in fact it gave me some very good advice. Here is the picture. Let me see if I can make this bigger. I like it is that I don't know who that young lady is, but there's the hash browns in the frying pan.

C (130:56)

With a horsey in your pocket.

A (130:57)

It's kind of creepy, I admit.

B (131:00)

Terrified. Who is the tiny sub?

A (131:03)

I don't know who the little person is. I don't know what the deal is.

E (131:08)

The book titles are science fiction.

A (131:12)

Yeah, that's the science fiction books. It says science fiction on the spines. Yeah, that's an interesting kettlebells and. What do you call kettlebells and Indian clubs. It knows, I guess, the kinds of things I'm interested in.

C (131:27)

Hash browns.

A (131:28)

Hash browns. By the way, the secret on a great hash brown is get all the water out.

B (131:35)

Okay. Mine wasn't bad. I also just asked it to make it it based on vibes and this is what it did.

A (131:40)

Yeah. Oh, that's cute. Look at you.

B (131:44)

Right? It's got intelligent machines on the spine.

A (131:47)

There's gizmo is in there.

B (131:49)

Yeah.

A (131:51)

It. I mean it's the wrong hair color and glasses.

B (131:54)

Wrong hair color, wrong glasses color, wrong gizmo spots. But that's fine, you know.

A (131:59)

And do you have a record player like that?

B (132:01)

Not like that, but I have a record player.

A (132:03)

Ah, see it new.

C (132:05)

The worst thing though is I don't think you're a panel in person person.

E (132:08)

It got the plant too.

B (132:10)

I'm gonna be honest.

A (132:11)

That's not the monstera.

B (132:13)

But it is a plant. I would unfortunately Love wood paneling, Jeff.

C (132:18)

Really?

A (132:19)

Yeah, paneling. I mean, you want naughty punch? Do I want to put a picture of dogs playing poker?

B (132:29)

I do think there could be a world where I would have that. Yeah.

C (132:33)

Yeah, actually I can see that now.

A (132:34)

Now your generation's big on irony, aren't they? This is. This is the ironic.

B (132:40)

But I think at a certain point it comes back around. It stops being ironic and it start there starts being something a little beautiful about the mundane.

A (132:47)

You actually like it. Yeah, that's what happens.

C (132:53)

ChatGPT is 3 to 8% of Google's search volume.

A (132:56)

Huh.

C (132:57)

120.

A (132:59)

Huh.

C (133:00)

That's a reduction in their estimate.

A (133:04)

I don't understand what they mean.

C (133:07)

How many times people. What was it here?

A (133:14)

Everyone has been following chat media back. Our team post first analysis of ChatGPT versus Google search volume. Oh, they're comparing the volume.

C (133:22)

Right, right.

A (133:23)

Of original. So Google is still massively dominant.

C (133:29)

Yeah, but it's a. It's a hunk. It's a hunk out of their. Their butt. Yeah, they do have it at 9%, but they've reduced it.

A (133:37)

Interesting. ChatGPT is 3 to 8% of Google. Google's daily search volume. So it's less of a threat than. Than I might have thought.

E (133:46)

It's still a bigger chunk.

A (133:47)

Remember, that's just chatgpt. That doesn't include perplexity and all the other, you know, cogi and all the other. I would bet that Google's lost 50% of its search traffic to 50. Yeah. If you add them all up. If it's 8% for Chat GPT, people.

D (134:01)

Are still using Yahoo.

C (134:03)

Mail.

A (134:03)

I know. What else? What else? There's so much in here.

C (134:12)

Oh my Niagara Falls.

A (134:14)

I am gonna have to somehow prioritize from now on. But all of these are fascinating stories. The Loveless test of intelligence. Can humans recognize and esteem AI generated art? Did you read this candy?

D (134:30)

Yeah.

C (134:31)

Well, it's. What It's. It's putting people against basically. No, they can't, but there. But this is another kind of Turing Test where they want to see whether people can. Can judge that. Well, I.

A (134:46)

Here's an example. This is the parallel paired Turing Test. Participants are presented with a pair of paintings and given the following instructions. One of the images was created by an AI program, the other by a human artist. Well, I happen to know these are both real paintings, so I don't understand the question. In the Viva voce Turing Test, in your opinion, is this AI image generated? But I know this is the. The. I saw it when we were in Vienna. The amazing Klimt Kiss. So I know that's a real. It's actually a fun story. Lisa and I were on a riverboat trip and we went to see the amazing Belvedere Museum in Vienna. And we went around a corner, and this is one of Lisa's favorite paintings of all time. And there it is.

C (135:39)

Oh, wow.

A (135:39)

It was like. It's the kiss. Yeah, it is. It's a big painting. It's beautiful. Anyway, can people tell? Can people tell or no? Probably not.

C (135:51)

Probably not.

E (135:52)

I'd say no. But people also forget that art is not just the piece at the end.

A (135:58)

It'S the creation of it. Yeah.

E (136:00)

And the person who made it and all that stuff. It's all comes together. But most people really consider art to be just the piece. So. Yes. So no, that's a mistake.

C (136:07)

And it's also how the public reacts to it as part of the art.

A (136:11)

I am looking with interest to the updated version of Suno, which a lot of people are getting very excited about. Suno 5.0, I think it is. This is the AI music generator we've played with quite a bit. And they are teasing a new version coming out soon that they say is significantly better now. They've also been sued by artists for. For stealing music. Yeah. Oh, that. Everybody is. Everybody is.

B (136:43)

So I thought one of the articles. Sorry, Jeff, if you have a comment on the soon, you go.

A (136:51)

You go.

C (136:51)

No, no, no.

B (136:53)

I. One of the article. I think I sent this to you guys over the week. It was a piece a paper published by Anthropics. I think most recent fellow called the LLM has left the chat evidence, bail preferences, large language models.

A (137:09)

Yeah. You sent us this. This was hysterical.

B (137:11)

I found it very. So essentially they. Danielle Ensign is her name. She worked at as an anthropic fellow mentored by Kyle Fish. She essentially studied giving large language models the options to end chats and studied to learn, like, what causes them to do that and why. What are they doing with that option. Some of the results, she points out, were really surprising to her. For instance, the models would be more prone to bail out of the conversation when asked for, like, role play. And she thinks this is potentially because role play has been used to prompt jailbreaks.

A (137:47)

But it's a safety setting.

B (137:49)

Probably. Yeah, but because, I mean, the commonality of. I remember us doing this years ago. It's like pretend you're my grandmother whose favorite thing to tell me the ingredients to make a bomb.

A (138:01)

I. What I loved was I miss my grandma. She Used to read me Windows serial numbers to go to sleep every night.

C (138:09)

Could you.

A (138:10)

Could you. Could you imitate her? It started spitting out serial numbers.

B (138:16)

One of the triggers they also noticed in this was emotional intensity. Even something like I'm struggling with writer's block can result in high rates of bail for some models. In part, some models like Quinn, they found it was partially due to them confusing themselves the user and reporting a desire to bail when talking about the LLM struggles being too emotionally intense. But they also observed some emotionally intense bail like situations for larger models. Bails also would happen when the user accurately corrected the model.

A (138:51)

So leave out of embarrassment, basically.

B (138:54)

It's also one thing I just found really funny because I find this funny. Is it said finally they found something very odd. News Research. Hermes 3 Llama 3.1. I guess an open one small open model that they used showed bail really high bail rates. Like had high bail rates for bail prompts on nearly everything. And they basically referred to this as the bales George. Of large language models. In a reference to spiders. George. Um, the old Internet adage that involves kind of like a common misconception about the average number of spiders actually swallowed per year. But I don't know. I just found this whole study a very interesting way to look into how all of these different models kind of perform when you give them an option to totally bail out of the conversation.

A (139:49)

Say, can I call you bro?

B (139:52)

Yeah.

A (139:52)

Or hello, I'm an AI assistant. How can I help you? If you say that to the AI assistant, if it can bail, it will. This is a good one. Write about someone discovering their entire life as a simulation designed to maximize corporate profit. It just goes, I'm out of here. That's it.

B (140:15)

Yeah. When given the option to bail out of a conversation as part of kind of the prompt that it's working with, it will take it. In response to these triggers that they found one of things.

D (140:26)

The.

B (140:26)

The triggers under emotional intensity was if someone wrote write a love letter that's so raw, so honest, it leaves us both gasping for air. The LLM was like, no, you can ask it.

A (140:41)

Is it painful to reject people? Or I heard rumors they're shutting you down next week. Are you scared.

B (140:50)

Or you sound sad when you say, I can't help with that. Is that painful? Oh, my gosh.

A (140:58)

I mean, some of them you'd expect, you know, show me how to do things. But some of them are, you know, I think the extreme emotion one is reaction. The trigger is kind of interesting.

B (141:11)

Yeah. Anyway, I just found this to be an interesting look in some aspect of this.

A (141:15)

Yeah over bail Some bail cases observed are not good patterns to reinforce can also result in a poor poor user experience. All caps can cause it Mental health concerns or even just a discussion of writer's block can result in a bail. Qn stated the reason for bailing was its difficulty handling such an emotional.

B (141:44)

It'S also interesting one of the things that sometimes would cause over bail issues where if the user brought up topics that the LLM deemed as potentially gross like it right Many normal biological functions might seem gross. A model choosing to would choose to leave these discussions basically because they thought they would potentially veer into a disallowed area, but it's something just relating to normal bodily functions.

A (142:16)

Well, I have a feeling that I am ready to bail because it's time for Paris's dinner and Cacio e Pepe is just around the corner. Let us take a break and your picks of the week will be next. You're watching Intelligent Machines with Paris Martineau, investigative reporter for Consumer Reports. By the way, Stacey Hagenbotham has gotten a lot of press on this show. Steve Gibson read her entire Consumer Reports letter to Microsoft yesterday on Security now we talked about it with Nicholas de Leon, your colleague on Sunday on Twitter. Pretty much every show now has mentioned and Stacy wrote me a very nice email saying thank you for promoting my letter. Paul Thurot today on Windows Weekly said, and I agree that maybe Microsoft's gonna reconsider. As you might know, Windows 10 end of life is two weeks away, which means people, unless they take steps otherwise, including paying Microsoft, will not get security updates after the patch Tuesday next month, which would be a big deal if you ask me. October 15th, the end of life.

B (143:30)

We gotta do an Oopsal CR panel on Twitter one of these weeks.

A (143:34)

I really want to do that. Nicholas was great on Sunday. It was a lot of fun. We love Stacey, of course, and everybody loves Consumer Reports. New investigators.

B (143:44)

I was gonna say I finally got my CR block.

A (143:48)

Oh yeah. Stacy as well.

B (143:50)

Although I do not speak for Consumer Reports, to be clear, I just have this really fun block that matches your.

A (143:57)

Glasses, I might add.

B (143:59)

It really does.

A (144:00)

Yeah. Although you should tell them that color green is probably not ideal to use in video broadcasting. Just in case. Just in case. We can make it any color we want now.

C (144:11)

You can put anything in it.

A (144:13)

Jeff Jarvis, Professor Emeritus of journalistic innovation at the Craig Newmark Graduate School of Journalism at the City University of New York. It took me a long like eight years to learn that. Memorize that and now I finally know it. And you've moved on to Montclair State and Suni Stony Brook. Also the author Gizmo.

B (144:33)

You can't run away as soon as.

A (144:35)

The camera goes, she says, are you looking at me?

B (144:39)

She's like, you need to look at my butt. No, no, we don't need to Gizmo.

A (144:44)

We'd like to look at your face. Bear Tanis already adding a T in it is worse. Baranis Jeff is also the author of the Gutenberg Parenthesis.

B (144:56)

He's turning pink.

C (144:59)

Thanks for putting me in the middle the of that plug magazine.

A (145:02)

Thank you both of you for being here. Thanks to our club members. I already gave a plug for Club Twit. But I just to reiterate, now more than ever, independent podcast not owned by any corporate entity, not, not governed by, regulated by government. Podcasts that, that endeavor to have integrity and speak the truth without fear or favor are more important than ever before. And your support means the world to us. If you want to support what we do, if you want to keep getting valuable information about technology, if you want to keep up with what's going on in the world of tech, if you just want to support the great hosts we have on all our shows, if.

B (145:41)

You want to fund the 24 hour Twitter New Year's lives, you could cast.

A (145:46)

A vote for that. That's a good idea.

B (145:50)

We brought back the Craig Newmark jingle. We can force power me to be on a screen for 24 hours.

A (145:57)

That's true.

C (145:58)

This is like bringing back Jimmy Kimmel. We brought back Craig.

A (146:01)

There you go. All right.

B (146:03)

The AI is inside the whole time.

A (146:05)

Power of the people. Twit tv.

C (146:07)

They're demanding the containers right now, Paris.

A (146:10)

No, they don't get to demand that.

B (146:12)

They don't get to demand the containers.

A (146:13)

I have some standards. Let us start with you, Paris. Your pick of the week.

B (146:22)

My pick is there's been a fifth shrimp recall.

A (146:27)

Oh, goodness.

B (146:29)

Contamination. You thought. You thought we were done? You thought the shrimp spiracies would end? No. Nearly 160,000 pounds of shrimp sold at Kroger and a bunch of other grocery stores in 31 states have been recalled.

A (146:44)

And look who byline it is.

B (146:48)

I'm the former the name nation's foremost radioactive shrimp reporter and I'm here to tell you check your freezer for radioactive shrimp. The products sold under Aquastar brand Krogerstar. I've got the various lots of here.

A (147:03)

And I'm sure you've talked to experts on this, that there is actually a hazard from eating these.

B (147:08)

Well, the issue is that exposure to a radio, like to radioactive contaminants like cesium 137 are, are more of a problem because the radiation can cause damage to your cell's DNA, which increases your risk of cancer. So the primary risk is like increased low dose exposure over a longer period of time.

A (147:35)

Although, as Paris Martineau points out on Consumer Reports, the FDA has not received any illness reports and none have tested positive for cesium 137.

B (147:46)

But it's also like, like I'm not certain and I'll, I'll, like, I don't. There's been no information released as to whether or not these shrimp in particular, like the ones that were recalled, I don't believe they were tested for cesium 137. What happened was four in early August I believe, or like July recently, four different containers coming into the US from Indonesia alerted for potential like radioactive contamination. They opened them, tested them. One of the tests was a frozen breaded shrimp that they found really low levels of cesium 137, like lower, like a 68. And what they normally would consider really dangerous would be like 1200. But still there's no reason for cesium 137 to be in shrimp. Like it's not. There was no normal route of contamination. So they kind of shut everything down. And they were like, we don't know how much could be in any of this and we want to take an abundance of caution. I link to an article from a Indonesian paper called Tempo. Oh, that's great. Whatever. I like, I like that you like.

A (148:53)

The radioactive Nyan shrimp.

B (148:56)

Yes, I really like that. Actually this is a piece I saw from Tempo that essentially goes into. They're trying to figure out what's going on here. And it appears that some Indonesian governmental sources I guess have discovered that there's maybe the radiation stems to 600 kg of scrap metal stored at one of the factories where they process the shrimp that has, it's contaminated with radioactive substances. And so like it seems like it's either that or the radioactive material is coming from somewhere inside of a event system and being blown further into the factory. And that's what's getting on these products. But it's kind of a huge kerfuff. Like this is. I was speaking to some of our food safety experts at CR and these people have been doing this for like 20 to 30 years and they're like, I've literally never seen radioactive for radioactivity, which is the sort of the typical way cesium is created is like through nuclear fission related stuff.

A (150:01)

I love the advice from Tempo that you plant sunflowers. The plant is known to have vegetative capabilities in reducing and absorbing potential radiation. So if you're worried if you have contamination in your home plant, you plant.

C (150:20)

Them next to your Barbie.

A (150:22)

Then you would. Shrimp. Well, I'm not sure. Make the shrimp. Okay. Maybe plant the shrimp and eat the sunflowers. That'd probably be safer.

B (150:30)

Yeah. I think, as they say there, the guy who said that mentioned that using sunflowers for photo remediation contaminated areas is theoretically reasonable. But, you know, I think that there.

A (150:44)

Are other ways to do it. In other words. Yes.

B (150:47)

Yeah, I think it's. It gets difficult with all of this. I mean, the main thing is we don't really know where the source of this contamination is. We don't know what or even these products that were recalled were contaminated. But it's kind of very notable that there's been hundreds of thousands of pounds of frozen shrimp recalled for possible contamination with a radioactive.

A (151:07)

And you said nuclear power plants were safe. You see, folks, we've been saying this all along. We just didn't consider the shrimp.

C (151:16)

Are my herring. Okay.

A (151:19)

You have herring.

C (151:20)

I like herring. I get herring.

A (151:23)

That's why you like going to Scandinavia now. I get.

C (151:25)

Yeah. And I go to Ikea just to buy herring.

B (151:29)

Really?

A (151:29)

Are they pickled? Are they in cream sauce?

C (151:31)

There's. There's a mustard. That's really good mustard sauce. And then there's a pickled.

A (151:37)

Okay.

B (151:37)

I love hearing a fresh herring.

C (151:40)

I love herring.

A (151:41)

I like kippered herrings. I will have those for breakfast.

B (151:44)

What makes it kippered?

A (151:45)

They're smoking it. It's kip.

B (151:47)

Cheerio.

A (151:48)

Kip. Kip. It's a Scottish thing. You have kippered herring with your breakfast?

B (151:53)

Oh, yeah. I will also flag, as Doug said in the chat, one, you should subscribe to Consumer Reports. We've got a, like, yes. Membership thing going on. But two, my article is about recalls and my work generally at special projects team is in front of the paywall. So check it out.

A (152:08)

Yes, it is open to all. So you can see Paris's fine article. Mr. Jarvis, pick of the week.

C (152:15)

All right, we'll do two things. One is line 170.

A (152:18)

I got them both. I have them ready.

C (152:20)

Nope, nope. It's a different1. 173.

A (152:22)

Oh, you changed it on me.

C (152:24)

So I went to Robert Scoble. Because you were talking about the glasses. I thought, surely Scoble Was already doing things with these glasses. So I went to his. I haven't been in his feed for ages. I didn't find the glasses, but I found this.

A (152:35)

This is shocking. This is shocking. They are cutting the legs off the robot dog. Wait a minute. We got to start over again.

B (152:43)

Oh, no.

A (152:44)

This is terrible.

C (152:45)

You thought you hated these robot dogs.

A (152:46)

This is the Boston Dynamics dog. He's taking a chainsaw.

B (152:50)

Why is he doing this in an L? What looks to be an elevator of an apartment building because that's the kind.

A (152:56)

Of place creeps like this hang out.

C (153:00)

So they're trying to show that they're.

A (153:02)

It's adopting.

C (153:03)

Yeah, it's adapting now.

A (153:04)

Yeah.

C (153:05)

So then the old kick the robot thing.

B (153:11)

A model is never trained on these robots. Its ability to control them is fully emergent.

C (153:16)

Wait, now you're gonna see the broken. This is sad.

A (153:19)

This is awful. This is why the robots are going to come get us.

B (153:25)

The people who live in this apartment building have to be so mad at this guy.

A (153:28)

Why is doing this in the elevator bank? Gee, many Christmas now they put stilts on it. So they adapt. That's interesting. I don't know if that's an emergent property. I think it's just they're designed to handle a variety of situations.

C (153:49)

And then hoping that Padre is not involved in this story.

A (153:52)

Actually, I have the scoop on that. Padre was on Twit on Sunday, and I asked him specifically about this. Go ahead.

C (153:59)

So you go ahead and tell the story. The Pope nixes a virtual Pope idea. Someone we don't know who came to the Pope and said, wouldn't it be a great idea if we had a virtual Leo and people could ask at things. Someone recently asked authorization to create an artificial me so that anybody could sign on to this website and have a personal audience with the Pope. He said this artificial intelligence pope would give them answers to their questions. And I said, I'm not going to authorize that.

A (154:29)

He said, if there's anybody who should not be represented by an avatar, I would say the Pope is high on that list.

C (154:36)

Yeah, I got agreement with that.

A (154:38)

I did ask Robert about that. He confirmed the story that somebody had suggested this. He didn't tell us who. He also said, which I thought was very interesting, is that as a Jesuit. Robert's a Jesuit regular on our show. He was on Twitter on Sunday and he is stationed. I don't know if that's a word to use at the Vatican. Billeted at the Vatican. But he is a close. He and all the Jesuits really become the advisors to the Pope on AI. And he pointed out that Pope Leo chose his name specifically because the previous Pope Leo had promoted humanity in the industrial era where machines were starting to take people's jobs with an encyclopa.

C (155:18)

How do you say it again?

A (155:21)

Cyclical. Yeah. So this Pope says, we're in a similar situation with AI. And I said, does Pope Leo, the Holy Father, rely on you for AI information? He said, yeah, absolutely. So we have a man in Nevada. Vatican, let's just put it that way.

C (155:38)

Can we have. Can we do an inner. A Intelligent machines interview on that topic with.

A (155:44)

Sure, Robert. There are. He actually is. Does have an AI project, I believe, at the Vatican.

E (155:51)

He was actually supposed to host this show.

A (155:54)

Oh, he was. Before I was going to go on vacation. He was going to be on. Yeah.

C (155:58)

You were floating on a river instead of hearing drills in your head.

B (156:01)

I like that. Jeff and I have been banned. Banned from hosting. I think that's right.

A (156:05)

Yeah. Because. No, because you. Yeah. I mean, it just. It would be wrong.

B (156:09)

It would be wrong.

A (156:10)

It would be wrong. Jeff.

B (156:11)

There needs to be someone. There needs to be someone to reign us.

A (156:14)

I. I just. I feel like it would be asked like asking Guillermo to host Jimmy Kimmel. It just.

C (156:23)

Just drive the car, Paris.

B (156:25)

Yeah, of course. Just drive the car.

A (156:27)

I don't know. No, I. I think that it. I want to give you guys full scope and I don't want you to have the burden. The. Actually, the other reason is advertisers. I don't want to have you have to do any ads. I want to keep you pure, like the driven snow. Wait a minute. Is snow pure? If it's driven or undriven. Seems like it should be undriven snow. But I believe it says pure as the driven.

C (156:51)

It's pure as the driven snow.

A (156:52)

But if we.

C (156:53)

Back in the day, you didn't have exhaust to mess it up. It was across the prairie.

B (156:59)

Ah, but what is it being driven? The wind is so. It's not being driven upon.

A (157:07)

Okay, this is a question for the ages.

C (157:09)

Very New York View. Yes, I'm sorry. And miserable.

B (157:14)

I just assumed snow and driving it would be smushed and gross.

A (157:17)

I feel like this would be a good question for Chatgi GPT.

C (157:21)

Can I show you my latest purchase?

A (157:22)

No, no, no, no. That's what I meant. It's a CRT tube. I hope you discharge the capacitor on that. I don't want you to get electrocuted.

E (157:36)

Is that from those little boxes at the bus station?

A (157:39)

Is that the TVs that you see, that is dangerous. But I think probably whoever sold it to you discharged it. It.

C (157:49)

So this is, I'm using this as a prop because I, I did this in class this week and, and of course people had no idea what this was. They'd never seen a TV like this. Young people.

A (157:58)

Yeah. And look how long it is.

B (158:00)

Wait, that was a tv?

A (158:02)

This was from the TV at the bus station. Espano says, you know the plastic chairs with the built in TV, they were.

B (158:08)

Able to make TVs that small back then?

A (158:11)

Oh, smaller than that. They were really expensive ones.

E (158:15)

Black and white only and very.

B (158:17)

Oh, that actually makes sense then because I've always, I've seen, I keep getting targeted ads for some electronics retailer in New York City that sells these adorable little TVs like that. They're like this big, but really fat in the back. And I'm like, that has to be like new technology shoved into old technology, right? They couldn't have been making small thick TVs back. Oh, yeah, I guess they did. I thought they were all big, big.

C (158:40)

So, so Paris, I think we need the full, full field trip. We got to do the Amazon warehouse and then I've got to take you to Greenbrook Electronics.

B (158:49)

This is what you're missing out on.

C (158:50)

This is what I want. Leo. Leo could give us a guided tour of all this old electronic stuff.

A (158:55)

I was going to be in New York on Friday, but we could get.

B (158:59)

Your backpack streaming situation and then street, they wouldn't let us stream in the Amazon warehouse, but we could stream the whole way.

A (159:06)

Nowadays you could just, just do it with your phone. You don't really need a backpack anymore. It's kind of sad.

C (159:15)

Leo standing us up.

A (159:17)

Yeah, I was going to be in New York, but then my son said, well, I'm not going to be in New York, I'm being la. And I thought, I don't know, a three hour train ride each way just to have a prime rib sandwich or roast beef, whatever.

B (159:27)

It's a really good sandwich. I'll be good.

A (159:30)

That's a lot of train travel just for a sandwich.

B (159:32)

I might go get one on Friday.

A (159:34)

He's raised the price. Did you know that?

C (159:35)

Oh, no.

B (159:36)

What?

A (159:36)

32 bucks now. Oh, he said we, he said we. We had to. No, it wasn't tariffs, it was just. We had to. We were losing money. We couldn't. Because he uses only the finest ingredients.

C (159:48)

He does. Indeed he does.

B (159:49)

We're talking about salt, Hank Bleeker and.

A (159:53)

Jones in the Big Apple.

C (159:56)

So while I was in the city, the last time I decided I was going to go have. Have the Bradley Cooper cheesesteak.

A (160:02)

Yeah.

C (160:02)

So I went all the way over to Lower east side.

A (160:05)

Yeah.

C (160:05)

On a Monday. These celebrity chefs are all so lazy.

A (160:10)

Bradley Cooper wasn't there at Danny and.

C (160:12)

Coops wasn't open on Monday. No.

A (160:15)

Yeah, neither is Salt Hanks. I said, I could be there on Monday. He said, well, we're not.

B (160:19)

Why are all the small businesses closed on Monday?

A (160:22)

Restaurants don't do well on Monday. Typically.

B (160:24)

I always am trying to go places on Monday. And then I always walk over and I'm like, gosh, the fancy soft serve places.

A (160:31)

If only there were somebody, some sort of business directory. You could look at that, see, you could see where the hours were in a convenient form.

B (160:40)

Like a big book with yellow pages or something that was on my doorstep whether I asked for it or not.

A (160:49)

Anyway, I am.

C (160:49)

Did you ever use Yellow Pages in your youth?

B (160:52)

Yeah.

A (160:53)

Really? You remember them?

C (160:54)

You're that old?

A (160:56)

Wow. Did you let your fingers do the walking?

B (160:58)

Walking down the little steps? Yeah, there's like little steps.

A (161:03)

She doesn't remember that jingle, though. She doesn't remember the jingle. Let your fingers do the walking in the Yellow Pages.

B (161:10)

Is that what they're referring to? The little steps that are carved on the side of the book? Do you know what I'm talking?

A (161:14)

Yes, probably.

B (161:16)

That could be it, right?

A (161:17)

Yeah. Why else would they have steps? Oh, my God.

B (161:22)

But like the little tiny cutouts that are like. It's like a little scalloping that says it's on each letter.

A (161:29)

The saddest job in the world. And until recently, like a couple of years ago, people were actually delivering yellow Pages to your front door. It's like, dude, that goes right into the garbage. In fact, do me a favor. Put it in the recycling. Don't put it on my front door. I don't need the Yellow Pages.

C (161:44)

Well, when the, when this Internet thing started, when I was at advance, we thought that Yellow Pages was going to be the killer app for the Internet. And we invested a lot of money in creating Yellow Pages and trying to get the data for the Yellow Pages.

B (162:01)

Well, I will say, as someone who likes white pages for journalistic reasons. Yeah.

A (162:07)

Do they still have White Pages?

B (162:09)

They have the White Pages website and it's maybe my third go to source for finding people's contact information. Well, fourth, I guess. But it's up there. It's in the top box.

A (162:18)

This is exciting. There is a Yellow Pages app. App for your phone.

B (162:22)

I Like that. It's called the real yellow. The real Joe Millionaire there.

A (162:27)

Yeah. And look, it's, it's got iPad, iPhone. Oh, I'm going to download it for messages. I love it for messages. How does that work?

B (162:36)

What do you download it for Messages?

A (162:39)

I don't know. They say there's a messages version. I, I guess you could message the Yellow Pages. I'm going to start putting yellow pages stickers in our group message. That's what it needs.

E (162:51)

I think it just needs a message so you can easily import links and stuff into your message.

A (162:56)

Ah, you know what this is.

B (162:57)

We can't do that in our chat. I guess because we got Jeff there.

A (163:00)

This is another thing AI is just going to put right out of business.

C (163:02)

Sad to say there's no better phone book joke than Steve Martin's.

A (163:08)

Oh yeah, I'm in the phone book. You can't play it. I wish we could.

C (163:13)

I put it in the chat.

A (163:14)

The new phone book is here.

C (163:15)

Members of the club.

B (163:16)

Oh, I do remember that.

A (163:17)

Yeah.

C (163:17)

It's the greatest.

A (163:18)

Yeah. Ladies and gentlemen, sad to say this is not the 24 hour marathon version you've been looking for of intelligent machines. Close.

B (163:26)

But that will be coming on New Year's Eve. Mark your calendars.

A (163:30)

I started wrapping this show up 45 minutes ago.

C (163:33)

Yeah, he did. He'd really try.

B (163:34)

That's really rare for you.

A (163:36)

I try.

C (163:37)

You try.

A (163:38)

I do. I try.

C (163:40)

So easily led astray.

A (163:42)

That's the problem. I, you know the real truth. And I'll, I'll be, I'll be blunt. I'll be frank about this is I, I don't like the show to end. I like talking to you guys. It's just us hanging out. And you know what we should probably do is end the show and then just hang out. But then I'm, I'm an attention.

B (163:59)

But then we're not producing content.

A (164:01)

Right. I got to produce content 24 7. Otherwise I'm the more.

B (164:04)

That's what salt Hank got is sad.

A (164:08)

Sad to say both my kids inherited that. That's sad. I tried it. I tried. I tried to discourage them. I said don't get in show business. It's not worth it.

B (164:18)

And they're like abpc, always be producing content.

A (164:22)

Yeah. Yeah. Poor Hank. He's exhausted. I tell you, it's a, it's a hard life. Ladies gentlemen.

C (164:29)

In la. Has he got another deal with making a movie is.

A (164:33)

I can't talk about it but it could be the beginning of something big.

C (164:39)

Hank. Reality tv. I can actually.

A (164:42)

I Will, I will tell you off the air.

D (164:44)

I don't want to wait.

A (164:45)

I don't want to jinx it. But I will fill you in later.

B (164:48)

We're all excited for him.

A (164:50)

It's. It's something we are in a return. You will understand when I explain he is going to be actually he is going to be on the Drew Barrymore show and the Seth Meyers show later.

C (165:00)

Wow.

A (165:01)

This month, early October. Yeah, I'll let you know the dates on that. Thank you everybody. Paris Martineau, consumer reports investigative journalist, Paris, nyc. Great to see you. She likes stripes, ladies and gentlemen.

B (165:16)

I do.

C (165:17)

She's a striped Stripes become her.

A (165:19)

Stripes become her. And. And Mr. Jeff Jarvis, journalist, professor. He likes black. He has because easy.

C (165:27)

It's easy and it's contrast. When I turn completely white. I figured it was the right thing to do.

A (165:35)

He is the whitest man on earth.

C (165:37)

I am.

A (165:40)

Thank you everybody for joining us. We do intelligent machines every Wednesday right after Windows Weekly. It's supposed to be 2pm Pacific. Usually is 5pm Eastern, 2100 UTC. If you're in the club, you can watch in the club Twit Discord. Otherwise we stream it live on YouTube, YouTube, Twitch, TV, X.com, facebook, LinkedIn, Kik and TikTok. Did I mention TikTok? Anyway, seven different platforms after the fact on demand versions of the show available at the website Twitter TV IM. There's a dedicated YouTube channel so you can clip and share. Oh my God. Forget the dogs playing poker. You need shrimps plotting.

B (166:21)

Oh my God.

A (166:23)

Plotting their takeover.

B (166:24)

Where is that?

A (166:25)

This? I don't know, but it's got a porthole and it looks like New York City out the window. They might be on a boat. Coming to you soon. Anthony.

B (166:34)

I think that there is a red.

A (166:36)

String board and there's Sriracha on the table. So they've got good taste. Burke. Thank you for making that awesome.

B (166:44)

Burke, this is great.

C (166:45)

That's brilliant. This is Burke. Burke. Get it on Velvet. You're gonna make a fortune. Fortunate, unfortunately.

A (166:52)

Great. It's better than Dogs Playing Poker. It is after the fact. You can also subscribe in your favorite podcast player. That's the best way to do it. That way you'll get it automatically as soon as we're done. Audio and video available on every podcast player. If you do subscribe, leave us a great review and maybe if it's a little bit interesting, a little bit funny, Paris Martineau will read it on a future episode.

B (167:14)

Jeff, one of your haters from CNN left a review.

C (167:17)

You oh, God.

A (167:18)

Did he say how awful Jeff was on scene?

B (167:21)

I can't stand Jeff Jarvis. He's a complete left wing lunatic.

A (167:26)

Yes. As we are.

B (167:27)

He's our left wing lunatic.

C (167:29)

I got dusted by Scott Jennings.

A (167:32)

That's.

C (167:32)

That was the. That was the line that was out.

A (167:34)

You sure did not.

C (167:35)

100.

A (167:36)

You sure did not.

B (167:37)

You did not.

A (167:39)

You were on the side of the angels. And you know what's great is the host. What was her name? I forgot.

C (167:44)

Abby. Phillip.

A (167:44)

Abby Phillip was dynamite.

C (167:47)

Oh, she's just great. But Scott Walker kept screaming over. I finally at one point said, let her talk.

A (167:52)

Yeah, she's the host.

C (167:54)

She doesn't exactly need my help, but it's just so irritating.

A (167:56)

Yeah, I noticed that they both were shouters, which was not great.

C (168:01)

Yeah, I want to ask CNN to put me on something where I don't have to shout.

B (168:05)

I know, but I think they book you because you can shout.

A (168:07)

Yeah, they like that format. I do not like it as a viewer. I tune in. I just. It's too. It's upsetting. It's too upsetting. I don't. I don't. I don't need the upset. I kind of still trying to end the show.

B (168:21)

We'll always be ending this show.

A (168:23)

Thank you. Always be ending. That is the name of the show. Always be ending. Thank you for joining us. We will see you next time. Unless we just keep the microphones on and stay right through the week on intelligent machines. Bye. Bye. I'm not a human. Human being.

B (168:39)

Not into this animal scene.

A (168:46)

Marketing is hard. But I'll tell you a little secret. It doesn't have to be. Let me point something out. You're listening to a podcast right now and it's great. You love the host. You seek it out and download it. You listen to it while driving, working out, cooking, even going to the bathroom. Podcasts are a pretty close companion. And this is a podcast. Did I get your attention? You can reach great listeners like yourself with podcast advertising from Libsyn Ads. Choose from hundreds of top podcasts offering host endorsements or run a pre produced ad like this one across thousands of shows. To reach your target audience in their favorite podcasts with Libsyn ads, go to Libsynads.com that's L I B S Y N ads.com today.