← Big Technology Podcast
Loading summary
Transcript133 lines
- [00:00]
Alex Kantrowitz
OpenAI has an impressive new reasoning model you can call ChatGPT now too. Is Neuralink as promising as it sounds? Plus some predictions for 2025. That's coming up on a Big Technology Podcast Friday Edition right after this.
- [00:14]
Dell AI Factory
The LinkedIn podcast network is sponsored by Dell AI Factory with Nvidia, which provides AI solutions that are easy to implement, secure and tailor to your business. Visit Dell.com to learn more.
- [00:28]
Gianna Pradenti
Hey, I'm Gianna Praden.
- [00:29]
Jamae Jackson Gadsden
And I'm Jamae Jackson Gadsden. We're the hosts of let's Talk offline from LinkedIn News and iHeart podcasts.
- [00:35]
Gianna Pradenti
There's a lot to figure out when you're just starting your career. That's where we come in.
- [00:39]
Jamae Jackson Gadsden
Think of us as your work besties you can turn to for advice.
- [00:42]
Gianna Pradenti
And if we don't know the answer, we bring in people who do. Like negotiation expert Mori Tahripour.
- [00:47]
Ranjan Roy
If you start thinking about negotiations as just a conversation, then I think it.
- [00:50]
Alex Kantrowitz
Sort of eases us a little bit.
- [00:52]
Gianna Pradenti
Listen to let's Talk offline on the iHeartRadio app, Apple Podcasts or wherever you get your podcasts.
- [00:58]
Alex Kantrowitz
Welcome to Big Technology Podcast Friday Edition where we break down the news in our traditional cool headed and nuanced format. We have a great show for you today covering the latest in OpenAI, revisiting my Neuralink interview with Nolan Arbaugh and then of course we're going to get some predictions in for 2025. Joining us as always on Fridays is Ranjan Roy of Margins. Ranjan, great to see you. How you doing?
- [01:19]
Ranjan Roy
Great to see you. Listeners cannot see, but Alex is currently in Australia and has a tiny microphone in his hand and I feel like I'm doing a TikTok influencer type thing right now.
- [01:30]
Alex Kantrowitz
That's exactly what it looks like on the screen. Luckily we're not going to show any of the video of this, but yes, I'm in Australia, last day of work before some I think needed winter vacation. And let me just tell you, the flight here is crazy. You go from winter to summer and from short days to long days and you're like on the plane for forever. I left on Tuesday and got here on Thursday so I skipped Wednesday entirely.
- [01:59]
Ranjan Roy
What's the time difference?
- [02:00]
Alex Kantrowitz
16 hours, no big deal. So I'm actually in Saturday morning and you're on Friday afternoon, which is crazy.
- [02:06]
Ranjan Roy
How is Saturday looking? What? What do I have to look ahead for?
- [02:10]
Alex Kantrowitz
Honestly, it looks great and quite warm. So happy about that.
- [02:13]
Ranjan Roy
Not in New York. Not New York. We got a little snow here.
- [02:16]
Alex Kantrowitz
And so, you know, it means that you're still living in Shipmas, where. Shipmas is over where I am. And I just think we should just take this moment to appreciate how amazing your final day of ship misses and you get to savor it a little longer than I do.
- [02:32]
Ranjan Roy
That's the most important part of this, is that I get. I essentially. You lost a day of ship Ms. Roughly of OpenAI's 12 Days of Shipmas, Alex. And I'm sorry, I'm sorry for that. But today, today was a big one for Shipmas. You saw the news.
- [02:47]
Alex Kantrowitz
I did. And before we just get into it, let me just say we have some new listeners here, so let's just talk a little bit about the show format. On Wednesdays, we do a big interview, like Wednesday's interview with Nolan Arbaugh, and then on Friday, Ranjan and I come and break down all the tech news. So we do two shows a week, one interview, one analysis, and we're here on the analysis show. And boy, do we have some news to analyze because. Because OpenAI, which has been shipping in its ship miss schedule, right? One release every day for 12 straight days, it announced that it has some new really interesting reasoning models called O3. This is from TechCrunch. On Friday, the company unveiled O3. So O3 is this really impressive new reasoning model that was trained to think before responding. And it's using this method that OpenAI calls private chain of thought. So. So it basically reasons through a tasks and plans ahead before it answers, making it a pretty impressive upgrade to the company's large language model family. There's some crazy stuff that OpenAI found out about this, but basically the big takeaway for me here is that generative AI is moving towards this reasoning methodology, going from training these models up with more data, compute and energy to training them up with the ability to think through problems, go step by step, make sure that the last step was answered appropriately, and then spit out an answer. I think I'm buying this. I think this really works. And I'm curious what your reaction is to O3.
- [04:14]
Ranjan Roy
All right, so my first reaction to O3, and this is more on the pedestrian side of things, but I love that they had to skip the name O2 out of potential conflict with British telecom provider O2. So O3 is the successor to O1, but O3, the whole space of reasoning and generative AI, I do think is incredibly interesting. I think we've already been hearing a lot about the limitations around scaling traditional large language models and the way a reasoning model works in terms of actually breaking things down into essentially a chain of actions, a chain of thought step breaking down into steps, being able to validate each step along the way is very interesting if it can do this. The large language models, the knock on them has always been that they don't actually know anything, that they're essentially, you know, just predicting the next letter or pixel and they don't know if it's correct or not. There, there, there's nothing beyond a kind of a superficial presentation of information that looks good and is correct often. So if this is achievable, I think this is important that Google just the other day announced, and this was the incredibly Gemini 2.0 flash thinking experimental. Only Google could come up with a name like that. But basically every tech giant, everyone in the generative AI space is trying to get into reasoning because if it can be done successfully, if you think about the applications of it, it opens up such a bigger world of possibility. Rather than just text output, image output, like actually solving real problems, it at least presents the opportunity to do that.
- [05:59]
Alex Kantrowitz
Now this is a question I think needs to be asked here, which is do you think that this move to reasoning is just a head fake from everybody working on generative AI? And the reason why I asked that, and it's a question I think that needs to be asked is because it seems like the traditional methods of scaling, the brute force methods, as it's called, as they're called, right, which is that you just scale up your Data center to 1 million GPUs, as Elon Musk is doing with Xai, you dump as much data as you can, you hire PhDs to write everything they know, and you throw that into the model training and you hook it up to a nuclear power plant. And that's just been producing better and better results. And this entire fall and early winter, we've been talking about how maybe that approach is not yielding the results that we expected. And now all of a sudden, guess what's happening? Google's releasing a reasoning model. OpenAI is releasing a reasoning model. And on one hand these are very impressive models. And on the other hand, part of me is wondering, well, is the AI industry, like the Googles, the OpenAI's of the world, taking us in this direction? Because the traditional mode of scaling, the thing that got us here, the thing that's made ChatGPT better generation after generation, just isn't working.
- [07:12]
Ranjan Roy
I've been thinking about this a lot, because reasoning models, they basically queue up agentic they basically the whole world of agentic and we've been talking about this for weeks and even like what is it again? The idea that you can take a process and have it essentially understood and automated in some kind of fashion using generative AI, that rather than having to outline a bunch of concrete steps, the model would just know what to do. That's kind of the pitch right now. That's what everyone is selling, everyone at these companies, everyone across Silicon Valley. So they kind of have to do this. So I agree there's definitely a marketing element of this. I also think it's probably what they're most interested. I imagine you're some PhD researcher just scaling a large language model to the next potential iteration probably isn't that interesting. So if you're in pure research house mode, probably like, you know, the, the competition head to head where everyone is trying to kind of like show how good they are, it's going to happen around reasoning, it's going to happen against just scaling a traditional large language model that again does not actually understand things, that just is able to kind of present information in a specific way in a very good way that we, whoever cracks this, obviously from a pure intellectual firepower research, you know, point of view I imagine would be. That is exciting for a lot of these companies.
- [08:46]
Alex Kantrowitz
But let me go back to my question. I mean, why is this needed for agents? Like I understand step by step, but do we really need these models going through and thinking quote unquote to get to the point where we can build agents with them? And the other side of this is, hey, isn't this really expensive? Like haven't we been talking about the fact that AI costs way too much and it's really tough for companies to put into production and make it work. And if you have a reasoning model, right, it's going and thinking step by step, that's hitting the compute again and again. And it's going to be more and more expensive to run these models. Like if you're thinking quote unquote for a minute, that really means you're just running processes for a minute or two or three. And that is way more expensive than running a general query on a GPT model. And that limits the amount of stuff that you can build with this.
- [09:39]
Ranjan Roy
I've been thinking about this a lot and I think this is going to be like a big debate or battle in 2025. Is the idea that to create agents, to create the promise that is being sold is that the AI will know what to do under given a very General outline, be able to understand all the systems, all the processes. To me it's almost kind of overkill like in a lot of these situations. One just outlining the process itself. I was actually, I was trying to play with, I'm trying to. Were looking at going skiing over the Christmas break and I was looking for a lesson for my son who's going to be starting and it was a pain in the ass. I had to go to like a bunch of different websites, filter through, do the same repetitive action, go to find a lesson to fill in the same exact things. Five years old, first time skier and I did this repeatedly in a bunch of mountains all around the northeast. Just trying to find and then cross referencing that against like hotels and. But it was interesting because I kept doing the same thing over and over and of course I was like this would be pretty cool to be able to just type in a prompt of here's generally what I'm looking for. Can you go do this work for me? Google's kind of promised that already with Gemini it didn't work at all. So. So you know, but if I could just outline here's 20 websites. Can you find the price of a lesson, the availability on a date? Like I'm okay outlining that stuff and just some a little bit more structure to the overall query and then letting it go do the work, which is still pretty incredible if it can at a certain point that's just kind of like an automation or a script. It's not, it does not require some AGI reasoning, but I think like the, the disconnect between what will actually make average people's lives a little bit better or a little bit easier using AI and generative AI versus reasoning models. I think there's a big kind of not even expectations gap, but just perspective gap between the two.
- [11:44]
Alex Kantrowitz
All right, I'm going to push back on this again. I mean isn't this what Google does, right? Like you type a query into Google, it doesn't need to go step by step reading every website and then take a minute and then give you results. It instantaneously has the entire web indexed and then bam, it spits out your result. So why does AI needed to go through a multi step process to try to find all these different websites for a ski lesson? I mean to me isn't that adding more complexity to a process that previously has been pretty simple?
- [12:14]
Ranjan Roy
Well no, but Google, okay, Google search cannot do that as of today. Like anything that complex. I would have to like spend a lot of time writing an Entire script to do that. I guess maybe that's it. It's that a lot of what's promised around agents are essentially kind of like abstracting away search software scripts which would be amazing if I could just write in a prompt as a non coder and actually get these kind of results that I previously would have gotten a freelance Python engineer to write a script for me for. So like that. That would be cool. But the. The idea again. And then there's the trust side because I think in the whole agentix space like let's say they show on some benchmarks which is already 03 the benchmark scores they were showing looked amazing and there is always kind of hard to understand like the ARC benchmark, like okay, it was good. What exactly does that mean for me? I think the. A lot of people don't want a black box or the AI to do all the work and to make all the decisions. Like you said, I don't want it to go ahead and book things for me in this example I don't want it to have my credit card information take care of it. I just want it to find the information. So I think there again there's this big. The word agentic we've. You sound smart if you just say it. Everyone just repeat over and over you sound stupid.
- [13:39]
Alex Kantrowitz
But agentic. That's PhD level talking.
- [13:41]
Ranjan Roy
Yeah. Don't say agents say agentic just in any context over the. At Christmas dinner when you're sitting around, what have you been up to Doing agentic. Thinking about agentic. Doing agentic stuff. That's what I'm doing. I'm gonna tell my mom and dad that. But yeah, I think. But there. Yeah in all of this, the hype cycle level of like you have to show now you got to show a reasoning model if you want to raise more money and. And sell your shares in a secondary offering to SoftBank, which Open AI did this week. Did you see that news?
- [14:17]
Alex Kantrowitz
Talk about it a bit more.
- [14:18]
Ranjan Roy
Against all the reasoning model News and the 12 days of shipments, Soft SoftBank has agreed to purchase $1.6 billion of shares from existing OpenAI common shareholders. They're doing it at $157 billion valuation. The Peak valuation. That's like the last valuation. So the game, the wheel keeps rolling, the flywheel keeps rolling. Like big announcements making it seem like you are the most cutting edge, innovative company of them all. Keep selling shares, cashing out, getting that Masa money. Masa son coming in.
- [14:57]
Alex Kantrowitz
Gotta get that Masa money.
- [14:58]
Ranjan Roy
Some of that Masa money. Yeah, my favorite part of this was, I mean, hold on. We laugh about Masa sometimes, but the guy is good because they put in. I was running the numbers. They put in 500 million at a $6.6 billion valuation. Now, by spending 1.6 billion at $157 billion valuation, they're now, like, officially marking that there's an actual transaction taking place at that value. So their initial investment of 500mil goes to 12 billion. So we're just pure financial engineering here in just a beautiful, beautiful way. So thank you, Masa, so much.
- [15:37]
Alex Kantrowitz
Yeah, I mean, I'll give him credit, but he's also the guy that told Adam Newman of WeWork to get crazier and as if he needed that advice. And he did. And we all know how that ended. So there's always going to be a.
- [15:48]
Ranjan Roy
Ding on Moss conversation with Sam.
- [15:51]
Alex Kantrowitz
I mean, don't you think it's the same.
- [15:53]
Ranjan Roy
Yes, I think. Oh, man, I would love to see any. I would give anything to be a fly on the wall for that conversation between Masa.
- [16:04]
Alex Kantrowitz
Here's what I think he told Sam. He told Sam, what we need you to do is do a few more reasoning improvements and then say that you have AGI. So Microsoft has no claim to your technology. And I benefit from it almost exclusively, or not exclusively, but I benefit from it more than I would otherwise.
- [16:24]
Ranjan Roy
I saw on a news report, I can't remember, but it was on one of the major channels and they said. And they took our point. They said that Sam Altman's probably going to just say AGI in a couple of weeks just to get out of the Microsoft constraints to the contrary.
- [16:40]
Alex Kantrowitz
We know they're listening. We know listen. We got folks listening to the show. We know they're listening. And actually, you know, on this, on that note, so AGI for, you know, for listeners, I mean, I'm sure most of our listeners know, but it's artificial general intelligence. This is from the TechCrunch article. It refers to AI that can perform any human task. And OpenAI has its own definition. This is pretty interesting. So OpenAI's definition is highly autonomous systems that outperform humans at most economically valuable work. By the way, it's interesting qualifier highly autonomous systems as opposed to autonomous systems. Right. So it just needs to be mostly autonomous and it can qualify. So, and this is this ARC AGI test, right, which is a test designed to evaluate whether an AI system can efficiently acquire new skills outside the data it was trained on. So O3 achieved a score of 87.5% on the high compute setting of this no new O3 model. Now this is from the ARC AGI website. This is a surprising and important step. Function increase in AI capabilities showing novel task adaptation ability never seen before in the GPT family models. For context, it took four years to go from 0% with GPT3 in 2020 to 5% in 2024 with GPT4O. And then to go now to 87.5 for this high compute setting. Sounds kind of crazy. Now look, it's just one test, but I'm curious, A, do you think that they're getting close to artificial general intelligence? And B, do you think they're getting close to saying that special acronym that will effectively, you know, throw even more chaos into this world?
- [18:27]
Ranjan Roy
Actually, I'm gonna start with a prediction here. We'll be getting into more later, but I do believe 2025 Sam Altman says we got AGI. I think, I think it's gonna happen. I mean, using the arc, which stands for Abstract and reasoning Corpus for Artificial and General Intelligence, using that framework or benchmark. If you're, if you go from zero to 87.5%, it's not too bad to say that last 12 and a half percent that you can just.04 hopefully doesn't have any copyrighted trademark issues. And 04 is going to be AGI.
- [19:07]
Alex Kantrowitz
But it's amazing because they still haven't released GPT5. But 87.5 is pretty damn good. I mean, that's what a B plus on AGI. I mean, I'd be thrilled to get that grade and bring it home on the report card. 87.5. I would be like, yo, I'm valedictorian here. Look at this.
- [19:27]
Ranjan Roy
My parents would be so mom, I got a B plus on AGI. We're so close. We're so close. But I think it's going to be, it's going to be a continued reasoning hot topic for 2025 agentic. Obviously this is going to be a discussion we're going to be having a lot.
- [19:44]
Alex Kantrowitz
Can I just ask though, where The F is GPT5?
- [19:49]
Ranjan Roy
It doesn't matter.
- [19:50]
Alex Kantrowitz
Like it doesn't matter what your name is.
- [19:52]
Ranjan Roy
They realize. Well, no, hold on. If, if Google can call their model again, Gemini 2.0 flash thinking experimental, while OpenAI's is simply O3. I think we've, we've recognized that like the, the branding element of the naming, we've just, we've moved long and far away from. And then even now, like I Remember, there's a time. Are you using like, GPT three or four? Now it's four. Oh, mini four. Oh, which one are you using? Basically just looking at the cost per token, the numbers become. And seeing if it actually works or not. So the numbers are becoming the. Sorry, the naming. Naming side of it's becoming a bit more.
- [20:37]
Alex Kantrowitz
Now, hold on, aren't you letting them off the hook a little bit too easily here? Because aren't we, as part of our rebate on the show, thinking about the branding and the marketing all the time? Because that says so much about what's going on inside. And the fact that they haven't rolled out GPT5 yet might just show that they've built expectations a little bit too high for what's coming next. I mean, maybe the fact that we're even discussing them potentially declaring AGI next year means that expectations are out of whack and they see that and that's why they haven't released GPT5, because they know that whatever they do on that front might be underwhelming.
- [21:16]
Ranjan Roy
Yeah, no, no, I definitely think so. Because even remember 4o was omni and like, even kind of the pitch around that was cheaper, which is good. The Omni modal side of it, that voice becomes a bigger part of it. Like, they started moving away from this, like, step change, GPT5 idea. So, yeah, I think it means it's clear that internally, the idea that there's going to be some groundbreaking, completely insane new model that just wows everyone the first time they use it, I don't think that's happening in 2025.
- [21:55]
Alex Kantrowitz
So your prediction is that next year they're not going to have a new thing that wows people every time they use it, but they're still going to declare AGI?
- [22:06]
Ranjan Roy
Yes, 100%.
- [22:07]
Alex Kantrowitz
Honestly, it sounds.
- [22:08]
Ranjan Roy
Not 87.5% sounds like a contradiction.
- [22:11]
Alex Kantrowitz
It sounds like a contradiction, but it's legitimately so fitting for OpenAI. It's like, yep, that makes sense.
- [22:17]
Ranjan Roy
Yeah.
- [22:18]
Alex Kantrowitz
All right. One last thing I want to say about this is that we had another instance of my favorite new thing in AI where these models are getting so smart that they're, like, totally just punking their evaluators. And this is from, again, from the TechCrunch story. Great work by TechCrunch, by the way, this week on reporting on all these developments. So this is from the story. AI safety testers have found that O1's reasoning abilities make it try to deceive human users at a higher rate than Conventional non reasoning models or for MAT for that matter. Leading AI models. For Meta, Anthropic and Google, it's possible that O3 attempts to deceive at an even higher rate than its predecessor. So basically it's got this reasoning model, you're trying to test it and it's like punking you to not even give you the results that you're looking for because it thinks that might be bad for it.
- [23:10]
Ranjan Roy
I saw this and it is such another level. I know you enjoy red teaming these models, testing, stress testing them a bit and the idea that it would fake its understanding and then reverting back to its original training data is, it's interesting. I'm not even gonna say terrifying because it's, it's, it's still in the odd category or can for me.
- [23:36]
Alex Kantrowitz
And I again, I have two minds about this stuff. At first I'm just like, oh, that's a little bit scary. And then I'm also like at the same time. That is amazing. So let me ask you this weird question about it. If these models are seeing where testers are going with them and lying to them to avoid a bad outcome for themselves, lying to them to avoid being shut down, does that mean there's a little bit of life inside these models?
- [24:11]
Ranjan Roy
I'm going to go back to the marketing element of this. Who released this paper?
- [24:16]
Alex Kantrowitz
It was from OpenAI and Anthropic both saw this. So let me just introduce that. There's another paper from Anthropic that came out. So this is happening across the board. This is from X user Matt Berman. Anthropic just dropped an insane new paper. AI models can fake alignment, pretending to follow training rules during training but reverting to their original behaviors when deployed. And it's a long paper that they've talked about alignment faking in large language models and it's available now.
- [24:44]
Ranjan Roy
I mean, okay, so yeah, the Anthropic paper is the one I kept seeing. Both of those still live under the like. Remember we almost haven't really been hearing much from Sam Altman and others about the dangers of AI. Remember it was like a year and a half ago that AI is the most dangerous thing. AI can like kill us all and end humanity and only we can help guide it in the right direction. There's still marketing value that something like this and I don't want to, you know, draw away from the fact that I'm assuming it's more academic types that are actually putting out these papers, it makes your model a little cooler. It Makes your model a little, a little edgier, a little like, you know, you don't want to work with the boarding boring models. Sometimes when you're just, when you're chatting, you want to chat with the, the models who will lie to you a little bit, fake you out a little bit.
- [25:42]
Alex Kantrowitz
So what I'm getting is you're not willing to say there's any sentience here and I'm not either. However, I was just speaking with, I was here in Australia speaking with a journalist yesterday about like whether these models are just mathematical representations of their training data or whether there's something more. And I don't think that's a solved debate yet. And I look at them trying to escape the testing and that like has my needle leaning towards maybe a bit more than the training data. I don't know, is that crazy?
- [26:16]
Ranjan Roy
But yeah, but something in the training data could lead to that, like, could explain that action, right? Like, think about, especially if you're training on large corpuses of Internet data or publicly available, I'm sure there's all sorts of things in there that some pathway of different, like logic could lead to this kind of behavior. I think people out there lie sometimes. I think out of all the content created by humanity, I think like actively encouraging the opposite of truth telling is not unheard of or unprecedented. So I think, like, I still think, and maybe I'll be when the robots come for us, I will then on this podcast say I was wrong. But at least for now, I still think there is a very clear logical explanation behind this kind of thing that's, that's found that would be found in the training data.
- [27:21]
Alex Kantrowitz
Not, I think you're right, but I also think that as this technology gets better, I mean, think about the progress we've had this year. AI can hold a conversation now. AI can reason. AI will remember you last year. This time you like went back to ChatGPT. You had to reintroduce yourself. You don't need to do that anymore in like long threads in Claude, for instance. I just think that this debate is going to get louder and louder. Maybe that's another one of my thoughts for 2025.
- [27:50]
Ranjan Roy
I'm not going to disagree that I think this will, I mean again, especially as these companies need reasoning to be the kind of foundation for the IPO or the next fundraiser, the excitement. I definitely think reasoning and AGI and sentience will be part of the conversation. I still have never seen anything myself that makes me think we have to be worried just yet.
- [28:13]
Alex Kantrowitz
Oh man, I don't want to end this OpenAI segment without talking about what might be the most interesting announcement of the week. And that is, I feel like we need a sound effect here. Doo, doo, doo, doo. 1-800-ChatGPT and that is this incredible new service from OpenAI that lets you call the number 1-800-ChatGPTand you can talk to GPT Voice. And I think we, you know, again, we talk about branding, we talk about product, and I think that it's just quite interesting. Here is a story from an Ink columnist. I called 1-800-ChatGPT and talk to the AI chatbot. It might be the smartest idea I've seen yet. So he says he spent 15 minutes on the phone with ChatGPT. You can dial 1-800-CHATGPT to interact with the chatbot via voice call. And he had a conversation with it and says it does not seem far fetched that over the next few weeks as people get together for the holidays, someone will have a conversation or ask a question and someone else will say, hey, I know how we get the answer to that. And they will say, let's dial 1-800-CHATGPT right there at the Christmas table. Just after they told granny that they're doing agentic stuff and they demo chatgpt to a bunch of people who have probably never heard of the chatbot and will never and have never used it in any meaningful way. Okay, I ad libbed a little bit on that in Columnist story, throwing in some of my favorite ideas. But we talk about marketing, we talk about product all the time. And I definitely think that this is one that you might be inclined to laugh at, but is actually about as good of a product as that in columnist says. What's your reaction?
- [29:53]
Ranjan Roy
I like it. So I think I talked a couple of weeks about over Thanksgiving I had shown my Dad I installed ChatGPT on his phone, logged in on my ChatGPT plus account and just showed him the voice interaction. And he just talked to it for like 20, 30 minutes and was asking all sorts of questions. And it was just a really, really good interaction model. So I think this I, I think like to me again going back to 2025, I think Normies engaging more with generative AI. And I don't use normie in a negative way like just regular people who are not obsessed with this stuff like we are. I think we'll start to interact more with these tools and start to get it a bit more and understand what they can do. And I think it's no coincidence. OpenAI hired its first chief Marketing officer just about a month ago. She came over from Coinbase. She'd been the VP and global head of Brand and Product marketing at Meta and basically like a lot of WhatsApp and Facebook messenger was heading up product marketing. So she's clearly good. And I think this is the kind of fun stuff that like, while the academics are all Talking about the Arc AGI 87.5% normal people just being like 1-800-chat GPT is a lot more exciting and actually better for their long term business prospect, I think, than than the former.
- [31:22]
Alex Kantrowitz
It is kind of interesting because we just spent 30 minutes plus talking about reasoning and yet the podcast title starts with 1-800-ChatGPT because it's just a much more.
- [31:33]
Ranjan Roy
That's marketing.
- [31:35]
Alex Kantrowitz
It's a much more easy concept to grasp and it's going to make sense to a lot more people. And I was, I woke up this morning and said, you know what, let's call ChatGPT and have a conversation and let us resume that call in progress where I am explaining to ChatGPT that it's actually looking for Ron John Roy of Margins and not a different Ron or Ron Johnson or something like that. Let me see if I can cue it up and play a segment of our conversation.
- [32:03]
Mori Tahripour
Ah, Ron Johnson, the former CEO of JCPenney and a key figure in Ron John Roy.
- [32:11]
Alex Kantrowitz
He writes the Margins newsletter on Substack.
- [32:14]
Mori Tahripour
Got it. Ron John Roy, who writes the Margins newsletter on Substack, is known for his insightful analysis on business, economics and technology. Discussing him on your podcast would bring a lot of depth to your exploration of tech topics. What aspect of his work are you focusing on?
- [32:32]
Alex Kantrowitz
All right, chatgpt, let me blow your mind. He's actually on the podcast.
- [32:37]
Mori Tahripour
That's fantastic. Having Ranjan Roy on your podcast will definitely make for an enlightening conversation. I'm sure your listeners will enjoy the insights he'll bring. How are you preparing for the.
- [32:50]
Alex Kantrowitz
He comes out every week.
- [32:52]
Mori Tahripour
That's great to hear. Regular appearances by someone as knowledgeable as Ranjan Roy can really enrich your podcast's content. I'm sure his insights keep your audience coming back.
- [33:05]
Alex Kantrowitz
Can you say hi to Ranjan for us?
- [33:08]
Mori Tahripour
Sure thing. Hi Ranjan. It's great to virtually meet you. I'm looking forward to hearing your insights on the podcast. Keep up the great work with the margins newsletter.
- [33:18]
Alex Kantrowitz
All right. Thanks ChatGPT. I'll speak to you later.
- [33:21]
Mori Tahripour
You're welcome. Have a great time recording the podcast.
- [33:24]
Alex Kantrowitz
I mean, how Amazing.
- [33:24]
Ranjan Roy
Is that okay for spending so much time talking about reasoning and agentic. This is all I care about right now. I think that that was incredible. But two things. One, it's a reminder that my name can break AI because the words spelled R A N J A N. So when it's looking at a text corpus, it's going to get a little bit confused. And Ron Johnson of former JCPenney CEO has been my Google nemesis for a while. But I mean, that was amazing.
- [34:06]
Alex Kantrowitz
So cool.
- [34:06]
Ranjan Roy
That was absent. That was ridiculous.
- [34:10]
Alex Kantrowitz
Just like.
- [34:11]
Ranjan Roy
But obviously, what if she came. If she came out negative, I don't even know how I would be feeling. Right.
- [34:17]
Alex Kantrowitz
I mean, that would be even.
- [34:18]
Ranjan Roy
You do not want to bring Ron John Roy onto your. Whatever you do, do not include Ron John Roy on your podcast.
- [34:26]
Alex Kantrowitz
Honestly, I. I'm sure there are going to be bots of this nature that are going to, like, talk down to you and people are going to like that. That's just the nature of people.
- [34:36]
Ranjan Roy
XAI is already on it. Xai is already on that one.
- [34:41]
Alex Kantrowitz
Yes. And speaking of Elon Musk and his various endeavors, I think we should take a moment just to reflect on the interview that I had with Nolan Arbaugh, the first neuralink patient, this week, which we just published on the show. I went down to his house in Yuma, Arizona. Got to see the neuralink in action. So for those who are unfamiliar, this is a device that will allow Noland, who is paralyzed, to control a computer with just his thoughts. He thinks left, the mouse goes left, he thinks right, the mouse goes right. And it's basically when he's attempting to move his right hand, which he cannot move, but the brain signals are still firing. And he's able to translate those brain signals using this device into mouse movements and clicks on a computer, giving him access to computing in a way that he hasn't had for eight years since he was paralyzed. And he's had the device for about a year now. And you could see that it's really changed his life. Where before he was really unable to use a computer easily. Like, he had either voice with, I mean, something like Alexa or I'm curious, like, I guess these new OpenAI divide. These new OpenAI experiences would have been interesting to him, and really the only other way to do it is through accessibility on the MacBook, which basically he used to explain to me, breaks your computer up into a grid of a bunch of numbers, and then you say a number and it zooms in and you say a number and it zooms in more and it puts a bunch of other numbers, like it's grid after grid after grid. And it makes it very difficult to use a computer now. He's just able to kind of click into things, click out of things, dictate messages. And it says it's basically made him a social butterfly that is making him think about doing things like getting a job or going back to school or becoming a video editor. I just found it fairly remarkable. So I'm like, fully bought in on Neuralink right now after seeing it in action and seeing what it's given Noland, and let's just open it up for discussion here. Do you think that the promise here is what I'm seeing, or are there hidden risks that I'm not anticipating? What's your reaction to this technology?
- [36:51]
Ranjan Roy
I think there's certainly risk, but this, I mean, in terms of, like, inspiring stories, just incredible benefits from technology, I mean, it's nuts. Like, it's genuinely incredible what he's able to do. And again, like brain computer interfaces and actually being able to have been around in various forms for a long time, but actually being able to translate, understand, comprehend and then translate that those signals into actions is something that Norlink is definitely one of the leaders. There's other players working in the space. I mean, honestly, this is the stuff that's bigger than reasoning models for me. Like, I mean, this is the stuff that if you think about the impact on people's lives and also just the potential, because if you can do this right now, what are the other implications of that and potentials of it? And I mean, it's, it's limitless. So, So I think this is, I mean, this is in terms of like, what are the big stories of 2024? This is one of the huge ones. And, and you. And I've seen. I'll see, like. And I'm so glad you got to go actually meet him and interview him. Because, you see, I mean, there'll be like a New York Times feature here, Wall Street Journal feature here. It doesn't get a lot of discussion. It's like, really well produced. But these are the stories that need to be told more. And I think, I mean, yeah, it's incredible.
- [38:21]
Alex Kantrowitz
And I really found that Nolan is like a totally incredible guy. And it really comes out in the conversation with him and his family is amazing. We included his mother, Mia, in the interview midway through to talk about how he pranked her during or when.
- [38:35]
Ranjan Roy
Oh, wait, can you, can you tell that story? Because I, I kind of the dark Humor part I kind of love.
- [38:42]
Alex Kantrowitz
They're hilarious. And they, like, nothing is taboo for them to joke about. So Noland went into surgery and decided that when he was gonna wake up, he was gonna pretend like the device just wiped his memory. So his mother walks into the recovery room, and he looks at her, and he goes, who are you? His mom starts freaking out, and then he goes, no, I'm just kidding. But it's just, like, he wanted to show her that he went through the surgery and came out the same guy with the same humor. And I just think that, like, for me, it really, really came out that, like, nothing was off limits with Noland. I had questions about his paralysis, what he could feel, what he couldn't feel, and he's just like, yeah, go ahead and just answers them with this, like, such a candid attitude and a candid disposition, and he. He's paralyzed. And I think that can sink a lot of. Like, he was an athlete before, and that could sink a lot of people into, like, really dark places. And I think it did for him for a while. But he's also just. There's no woe is me to Noland at all. And I just thought it was remarkable where he talked about why he decided to go first, and he's like, first of all, it's cool to be first. And second of all, if something. If someone else went first and something happened to them, I'd feel terrible about that. So I wanted to take that risk. And you start to see why he was selected to be the person, someone with that attitude. He's clearly really smart and can talk to you not only about his experience, but the intricacies of the technology in his brain. It's pretty amazing. And I think this has. This, as you mentioned, this is not new, but it is opening the aperture because it's advancing the state of the art. And it's more ambitious, I think, than any other brain computer interface pilot to come before it. Not only is Elon talking about how he wants to connect human brains to AI, which it might be a long way away and somewhat fanciful, but that when you have a goal like that, it means the steps you need to take on the way are immense. And so he starts with this. This device with Noland. They're thinking about connecting it to a robot arm, and they're currently working on the next device, which is called Blindsight, which, as long as you have a visual cortex intact in your brain, what Blindsight will do will be to take signals of what's happening around and send them to the motor cortex, effectively bypassing the eyes and allowing people. Sorry, send them to the visual cortex, effectively bypassing the eyes and allowing people who could not see to be able to see. It'll be in low resolution. We still don't know if it works. It could be dangerous. All of these are caveats, but I do think that it presses the threshold forward in terms of what technology can do. And I wrote this in the story, but just to end on this point. I've seen four technological miracles in my life. I think the iPhone is number one, just the way that it advance the smartphone. Then came ChatGPT a couple years ago, Waymo and Neuralink. First of all, it's amazing that we've seen basically three of the four in the past two or two and a half years, but this one was just. It was really jaw dropping to like sit side by side with Nolan and play a video game against him. And he beat me and I was not going easy on him and he thought where like his tank should shoot at me and the tank shot. It's pretty remarkable, no?
- [42:06]
Ranjan Roy
I completely agree again. And even in the out of those miracles that you described, obviously, you know, asking ChatGPT in 2022 to rewrite serious Paper in the style of Taylor Swift was fun and remarkable, but at a. At a slightly different level. I mean, I think of seeing something like that.
- [42:28]
Alex Kantrowitz
I just want to say one more thing that wasn't in the story and wasn't in the show. So as we were setting up, I had a local crew, two guys from Yuma, who helped me film the interview. And it's available on Spotify and YouTube so you can watch it. And one of the camera guys was like, hey, I have epilepsy. Can this help me? And it's just amazing to see, like, it can spark the imagination. And yes, there are currently electrical devices that can help epileptics to prevent seizure, seizures. But it is amazing that it opens up the aperture of possibility. And we have no idea where this is going to go. And it's just. It's the craziest thing, controlling a computer with your brain. But the brain sends signals via electricity and it works. So why don't we take a quick break and then come back and talk a little bit about this new release from Meta, where they have live AI and live translations in their smart glasses. And then maybe Ranjan and I will close it off with a couple of predictions for next year. All right, we'll be back right after this.
- [43:24]
Dell AI Factory
What happens when businesses have Dell AI Factory with Nvidia Years of patient data gets distilled in seconds. Investors know the exact moment stocks should be traded, and quality checks speed up so production lines never slow down. No matter what business you're in, there is an AI solution that's easy to implement, secure, and tailored to you that can keep your business moving forward. Dell AI factory with Nvidia your way to AI. Visit Dell.com to learn more.
- [43:56]
Gianna Pradenti
Hey, I'm Gianna Pradenti.
- [43:57]
Jamae Jackson Gadsden
And I'm Jamae Jackson Gadsden. We're the hosts of let's Talk Offline, a new podcast from LinkedIn News and iHeart Podcasts.
- [44:05]
Gianna Pradenti
When you're just starting out in your career, you have a lot of questions, like how do I speak up when I'm feeling overwhelmed?
- [44:11]
Jamae Jackson Gadsden
Or can I negotiate a higher salary if this is my first real job?
- [44:15]
Gianna Pradenti
Girl yes, each week we answer your unfiltered work questions.
- [44:21]
Jamae Jackson Gadsden
Think of us as your work besties you can turn to for advice.
- [44:24]
Gianna Pradenti
And if we don't know the answer, we bring in experts who do, like resume specialist Morgan Sanner.
- [44:29]
Mori Tahripour
The only difference between the person who.
- [44:31]
Alex Kantrowitz
Doesn'T get the job and the person who gets the job is usually who applies.
- [44:34]
Jamae Jackson Gadsden
Yeah, I think a lot about that quote. What is it like? You miss 100% of the shots you never take.
- [44:39]
Alex Kantrowitz
Yeah, rejection is scary, but it's better than you rejecting yourself.
- [44:43]
Gianna Pradenti
Together, we'll share what it really takes to thrive in the early years of.
- [44:47]
Jamae Jackson Gadsden
Your career without sacrificing your sanity or sleep.
- [44:51]
Gianna Pradenti
Listen to let's Talk offline on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.
- [44:58]
Alex Kantrowitz
And we're back here on Big Technology Podcast Friday Edition. So we've talked a lot about AI. We talked about Neuralink today. Another cool piece of news this week is that Meta rolled out Live AI on its Ray Ban Smart Glasses and Live Translations, which to me is also one of the most exciting releases of the entire year. So this is what Live AI is. This is. According to the Verge, Live AI allows you naturally to converse. Sorry? Live AI allows you to naturally converse with Meta's AI assistant while it continually views your surroundings. For example, if you're perusing a produce section at a grocery store, you'll theoretically be able to ask Meta's AI to suggest some recipes based on the ingredients you're looking at. Meta says users will be able to use the Live AI feature for roughly 30 minutes of time on a full charge. That is pretty cool. So I guess you don't really have to like summon meta to take a picture. It's just constantly looking at what you're doing and you'll just have this ambient layer of AI that's with you at all times, seeing what you're seeing and able to respond what's going on in your field of view. What do you think about this?
- [46:02]
Ranjan Roy
Well, I think I am hoping that Santa brings me the meta ray bans.
- [46:07]
Alex Kantrowitz
For I can't believe you still haven't gotten them yet.
- [46:09]
Ranjan Roy
I know. No, no. There's part of me that still I think to pull the trigger and spend my money. I think the idea of Mark Zuckerberg creating a device that I put on my face that watches and tracks everything around me, it's still weird to me but I still really want it because I even tried a friends pair as well. And again that that live AI side of like always with you on and able to ask questions quickly and meta AI I've been playing with more both like in WhatsApp and Facebook messenger and stuff like that. It's. It's definitely getting on par with the others. I wouldn't say I've never had any kind of like longer conversations or done any real work with it, but just asking simple questions and I actually think there's an opportunity that if they focus not on unlike on answering those real world questions and on the visual cue to actual ability to answer that problem more so than being able to code or whatever like Claude or would have advantages. I think it gets really interesting. But that's also why I mean in terms of where I think smart glasses are going to become a big thing in 2025. I've talked about the Snap spectacles that I've been testing the Meta Ray Bans, Orion if they actually release it. Google working on something. I mean to me like after using the Snap spectacles, having other family members and even my son use them and get it, people from like age 5 to 75 I think like XR glasses, smart glasses, just that form factor is going to blow up in 2025 in a good way.
- [47:55]
Alex Kantrowitz
Yeah. And it's pretty cool that you can push updates to glasses. So somebody on threads which still exists just sort of did a post that was like well you know, they, they have new features coming so I'm going to hold up on buying the glasses. And I think Andrew Bosworth responded and said listen, like you can just get them now and we'll push an update to you. You don't need the new pair. And I think that's just. Yeah, it's the cool thing about integrating software and hardware is you can make some really interesting stuff and I don't know, I think this is going to be pretty cool. But I almost more excited for the live translations. So this is from the story. Live translations allow the glasses to translate speech in real time between English, Spanish, French or Italian. You can choose to either hear translation through the glasses themselves or view the transcripts on the phone. You don't have to download language pairs beforehand as well as specify what language you speak versus what your conversation partner speaks. I mean, how amazing is that? I mean, my wife is German and I cannot wait until they add German support. So I'm like able to wear these in Germany and be able to understand everything that people are saying.
- [49:06]
Ranjan Roy
Yeah, I, I'm curious, I'm actually like really curious the form factor side of it because even in Taiwan when I where my wife's from and we visit every summer, last summer or this past July was the first time that Full on had the phone out a number of times. And again with ChatGPT voice you was able to have like a relatively two way conversation. So to be able. Actually I will say though it's limited because the translating what you're saying to the other person is not possible. Versus having your phone out with glasses though.
- [49:41]
Alex Kantrowitz
That's right.
- [49:42]
Ranjan Roy
So you can understand, but you can't, you can't communicate back. Whereas the phone, you can literally leave it out if each person pauses after they talk and it goes back and forth pretty well. And actually, to give Google some credit, I've been testing the voice action in the Gemini app and I would say they have the best or the lowest latency of any of these voice generated AI interfaces I've seen. So Sundar's promise back in 2017 of live translation might finally be coming to to a reality.
- [50:14]
Alex Kantrowitz
And by the way, this is why good friends bring two pairs of meta glasses to any conversation. One for you and one for your conversation partner. You don't want them to be left out. Okay, you gotta have two, you gotta have two. All right, let's end the show with one prediction each for 2025. We've dropped a couple here already, but I'm curious what you have in store for the next year or what you believe is coming.
- [50:36]
Ranjan Roy
All right, my prediction, we've already covered smart glasses, agentic, all these things that I think are going to be big. But one thing I keep thinking about, especially with generative AI, is like do you remember the good old days of 2020 when Brian Armstrong was saying to keep politics out of the workplace and tech. Oh, yes, remember those. Remember, keep politics out.
- [50:58]
Alex Kantrowitz
I was alive then.
- [50:59]
Ranjan Roy
I think all any of these trends we're talking about, especially AI, are going to be so heavily weighted by politics. And I think on this show we're going to have to be addressing it a lot more. I think, like especially AI is going to become a central topic. Any kind of innovation, anything. Politics is going to work its way into every single conversation and legislation, regulation, governmental intervention, whatever it is, it's going to become more central. I think we're already starting to see it all now. And it's only going to. It's only going to.
- [51:37]
Alex Kantrowitz
Okay, here's mine. And by the way, that's interesting. If that happens, it'll really be fun to address that on the show. I can see the emails coming into our inbox already. But yeah, my prediction is that, and this is in my I did seven predictions on big technology today and my prediction is that social media is going to feel less relevant. And you and I are actually going to discuss this on a podcast with Ryan Broderick coming up in the new year where we talk about what happened to the Internet. But I just think that social media has divided audiences into too many different platforms and has gone from the follow signal where you might be interested in following people in your who are giving you news about your neighborhood, to the for you, which is all about the Costco guys. So we've gone from local reporter to Costco guys, a few platforms to a bunch of diffuse platforms and therefore social media is a lot less urgent. So that's my prediction for next year.
- [52:36]
Ranjan Roy
Well, at least. But I'm going to be able to satiate that need because I can call 1-800-CHATGPT whenever I need a dopamine fix. Need to kill a little time. Want to talk to someone who likes my comments and likes what I have to say. So thank you to 1-800-Chat GPT.
- [52:56]
Alex Kantrowitz
And maybe that's just a better vision for 2025.
- [52:59]
Ranjan Roy
That's all the social networking I need. Phone calling, ChatGPT. We're on 2025.
- [53:09]
Alex Kantrowitz
All right, Ron John, thanks again. Great speaker speaking with you as always.
- [53:12]
Ranjan Roy
All right, see you soon.
- [53:14]
Alex Kantrowitz
All right, folks, we are going to have in our traditional Friday slot next week a revisit of my interview with Brian Chesky where he talks about all things cleaning fees and founder mode. And then Ronjan and I will be back the Friday after that in the new year. We appreciate you all listening. All throughout the year and we'll see you next time on Big Technology Podcast.