Loading summary
Transcript174 lines
- [00:00]
A
How can you grow your business from idea to industry leader? Bring your vision to life with smart business buying tools and technology from Amazon Business. From fast free shipping to in depth buying insights and automated purchase approvals, they deliver everything you need to achieve your goals. It's not easy to stand out from the crowd. Simplify how you stock up to get ahead. Go to amazonbusiness.com for support. Hiscox Small Business Insurance Knows there is no business like your business. Across America, over 600,000 small businesses, from accountants and architects to photographers and yoga instructors, look to Hiscox Insurance for protection. Find flexible coverage that adapts to the needs of your small business with a fast, easy online'@hiscox.com that's his co x.com there's no business like small business. Hiscox Small Business Insurance so have you.
- [00:57]
B
Heard the story about the prescription plan? With savings automatically built in, it's where a family of any size can feel confident the cost of their medication won't hold them back. Go to CMK Co Stories to learn how CVS Caremark helps members save just by being members. That's CMK Co Stories.
- [01:23]
A
Bloomberg Audio Studios Podcasts Radio News.
- [01:39]
C
Hello and welcome to another episode of the Odd Lots podcast. I'm Joe Eisenthal.
- [01:43]
D
And I'm Tracy Alloway.
- [01:44]
C
Tracy, have you played around with GPT5 much?
- [01:48]
D
Not really. I've been perplexity pills.
- [01:51]
C
Oh, that's your main.
- [01:52]
D
Yeah, that's my main one at the moment. But is it good?
- [01:55]
C
I hear mixed so I use it because I use ChatGPT every day. It does not strike me as like obviously better for my uses than like the O3 models, which I've been very impressed by because you know, I want to establish I'm no hater or anything like that. But like it did not strike me as like oh this is like an amazing. Yeah, this is the thing, step function or whatever.
- [02:15]
D
It feels like the sort of breakthroughs, awe inspiring breakthroughs are kind of behind us and a lot of the progress on the models feels very incremental at this point, even though people are spending a lot of time and resources on doing it.
- [02:28]
C
The one thing GPT5 does prompt me and say, oh, that's a great question. Would you like to follow up more on that?
- [02:33]
D
But it's like, does it say oh Joe, you're so smart, that's such a smart question.
- [02:37]
C
You know what it did say? I asked a follow up and it started an answer with love it and then said love it. Do you want me to look into that.
- [02:44]
D
Yes, they are very flattering, aren't they? Actually, that's one thing I like about perplexity is it doesn't really flatter you, it just spits out an answer.
- [02:52]
C
So anyway, there's so many questions I have about AI and we talk about the business a fair amount and Nvidia and all that stuff. We actually don't really talk that much about the pure research side as much. But it's pretty important, I think, because I think a lot of people would agree that if the skills are like slowing down or if there were a wall or something like that, that might change some of these business model calculations, et cetera. So I think it's good. We need to get an update on just sort of the state of the art the science of AI.
- [03:18]
D
Yeah. Also it would be nice just to understand what's possible, possible in terms of the AI models and what people are actually researching, what they're working towards. Like, is it mostly about price, is it mostly about the output, Is it mostly about energy use? All those things.
- [03:33]
C
All those things. Well, I'm really excited to say we have the perfect guest, someone who is an AI researcher. We're going to be speaking with Jack Morris. He's currently about to finish his PhD at Cornell in AI. He's been affiliated with Meta professionally, so presumably he already has a $100 million pay package in the bank. No, he's shaking his head. He's not. That's a joke. But Jack, thank you so much for coming on Odd lots.
- [03:57]
E
Yeah, thanks for having me. This is going to be fun.
- [03:59]
C
What do you explain to me like what you're up to? Because I don't really understand how it works where people are. They're at a university and they're also at a company. And this isn't how it works in much of the world. Right. People get their degree and then they get a job. I get the impression that in the AI world it's a little fuzzier in terms of one's affiliations between industry and education, stuff like that.
- [04:19]
E
Yeah, that's definitely true. I think it might be on the way out, but I can tell you about my situation. So there's kind of a public research world and like a private research world. Okay. And all the academic institutions do public research and the AI labs like OpenAI, Anthropic, Google, DeepMind, they essentially do private research where they have these people in house that are running experiments and learning more about their systems, but they don't publish anything or share any of their knowledge. And so a cool thing about getting your PhD right now is you can do research, write about it, and then publicize it, like put it online. I tweet about it. I kind of like, can talk to you about it. And there's a few places left that'll.
- [05:01]
C
Still kind of the moment you go in house, we're never going to hear from you again. Severance or something like that.
- [05:05]
E
Yeah, I'll make sure they have a clause in my contract that I can still talk to Joe and Tracy.
- [05:10]
C
Okay, good.
- [05:10]
D
The odd lots clause. Yes, that would be important. So when we say AI research or an AI researcher, what exactly does that entail? Can't the AI models just research themselves? Just let them do it?
- [05:23]
E
Yeah, that's actually a very smart idea. And like, people are really worried about that, actually. Like, if we get to the point where the AI can improve itself and do research on. Yeah, then it sort of gets smarter and then it improves itself again and it ends up being this kind of exponential improvement that ends up with all of our demise. But I think right now it's not quite there yet. Like, maybe you can talk to ChatGPT.
- [05:47]
D
That's good news, Joe.
- [05:49]
E
Yeah, and good news for me too, because it means I can still get a degree and be gainfully employed, but I think it's still helpful. But we still need, like, humans to make these improvements. And in terms of what the actual day to day work looks like, I think it really varies. Like there are some people working on trying to make the models run faster or trying to make the hardware that runs the models run faster, more efficiently. There's people that try to work on the data, like, what should we train on? More coding problems or more textbooks or more Reddit posts, What works best to make the models? And then there's a lot more people working on different areas of the stack, like training algorithms. I kind of have my own little niche.
- [06:27]
C
What's your niche?
- [06:29]
E
There's this old field of information theory from like the 20th century where they talk about bits, like a zero or a one is a bit and you can add them up and have kilobytes and megabytes. And so I've been trying to think about what that means in like the ChatGPT world. If you train a model on a certain number of bits, how many bits does it actually learn? And like, can you look at the model and figure out, like, if you have one slice of the model, how many bits that is and stuff like that? So maybe the easiest way to explain is if you had for some godforsaken reason to use ChatGPT as like a flash drive like you, you had a certain set of data and it had to memorize all that data. Like how much data could it actually store? Store? That's the kind of area I've been working in. And then once you're there, you kind of realize, we could do this, or maybe next semester if we have time, we could try this other thing. And so it kind of branches out and there's a lot of little problems that you can try.
- [07:19]
C
I mentioned GPT5, it seems fine to me. It does not strike me as like, oh, because actually, so the first time I used ChatGPT, I was genuinely blown away like most people. And then actually I was pretty blown away by the O3 models in part because of how well they could do document search. And it's superior to Google search in many respects. And also just the organization of a lot of unstructured data, et cetera. Like, I didn't have like some oh my God, wow. Moment of. With GPT5, it's like, it seems like, how do we measure whether AI is getting better all the time?
- [07:53]
E
Yeah, that's a, that's really a huge question.
- [07:56]
C
Well, let me ask you. Okay, let me ask you actually a more specific question. How do the entities that test AI models as their job or as their function? What is the formal testing process look like to rank the quality of AI models?
- [08:13]
E
Okay, yeah, that's more tractable. We can start there and then we can talk about O3 and GPT5. So there's essentially two ways people do this kind of model evaluation. The main one is just by testing them on different data sets. So for example, there's this dataset called swebench that's a bunch of software engineering related coding problems. And they all have a human written solution and tests. And so you can ask GPT5, can you write the code for this and then run the tests and see if it's right. And still the models are pretty bad at that. I think they can do about half of them. They're very hard. They're like entire days of work for professional software engineers. But when a new model comes out, they can say, oh, look, we actually got a higher score on suite benchmark. And there's a ton of different data sets like that. So when GPT5 comes out, they say it's better at these types of coding tests. And a big one that specifically OpenAI has been advocating for is math. Like they did the International Math Olympiad and they said essentially GPT5 scored at the level of the best high school mathematicians, which is pretty cool. But you raise a good question of how does that actually map to real world usage? And, and I think this is like a really hard problem that people still haven't figured out.
- [09:28]
D
Does anyone try to capture that sort of like je ne sais quoi, I guess, when it comes to AI models, is one of the tests asking it to, I don't know, come up with a stupid limerick or something?
- [09:39]
E
Yeah, there are a lot of tests like that. There's some creative writing benchmarks and some poetry related ones. But I think you pointed out something interesting that for example, I mostly use Claude for manthropic and I think Claude does have this something to it that's like a little bit different and it's very difficult to characterize. It's just sort of the way it speaks to you and the way it thinks of itself is. I like it a lot better. But I don't know how you would design like a data set that can really capture that. The second way they do the evaluation is by. They call it, it's ELO scores, like in chess. So they, for example, ask the two models to write a limerick and then they have humans rank which one is better and they make this kind of ladder of Elo rankings for models. So I think right now Claude or GPT5 or maybe the Google model is top on this Elo ladder.
- [10:31]
C
The algorithm made famous in the social network that Mark Zuckerberg used to rate the hotness of his colleagues, still the workhorse model for comp evaluation.
- [10:40]
D
That's some good trivia, Joe. Very good.
- [10:42]
E
And no comment.
- [10:43]
D
Well, I assume just on the hard number, evaluation people are also ranking these on data usage, energy, that sort of thing as well, right? Speed. Speed would be an obvious.
- [10:56]
E
Definitely. The AI companies like to use price as a metric, which is kind of interesting because there's a lot that goes on behind the scenes, including just sort of like free money that drives the prices down. But they also do benchmark speed. And I think you make a good point that the benchmarks can be pretty misleading. Like for example, there's a bunch of recent open source models that came from different Chinese AI labs that have really, really high scores on certain benchmarks. But people kind of think they're not as good for real world usage.
- [11:25]
C
For whatever reason I've seen people talk about this. Isn't part of the problem with testing AI or evaluating AI that a lot of these problems exist in the real world already? Right. You see this a lot. And people are always finding this, which is that here's an AI model that is amazing at math on the Math Olympiad. And yet it gets tripped up by questions like which is heavier, a pound of steel or two pounds of feathers. And they'll say that's a trick question. A pound of steel weighs the same as two pounds of feathers. Which is clearly like, it was clearly then been trained in some sense to recognize these steel versus feathers thing or whatever it is. I forget if it's deal, but it also clearly can't measure whether one or two is bigger.
- [12:11]
E
Yeah, that's a really good example. I think they kind of successively include these kinds of things in more rounds of training data. And so every time a new model comes out, they kind of patch little holes that appeared in the previous models. So you're pointing to this like they probably started with the classic riddle, that's like a pound of bricks or a pound of feathers. Bricks and they're equal. But then like, the models got that wrong. And so they add it to some.
- [12:35]
C
Part of the training, like a very efficient way to achieve intelligence. Like, oh yeah, we should have included that. Oh yeah, we got to include that. Oh yeah, we got to have.
- [12:43]
E
Right.
- [12:43]
D
Like every little thing, like going.
- [12:46]
C
That does not speak to me of a line towards something that we would call anything resembling human intelligence.
- [12:53]
E
I definitely agree. I. I think one counterexample is people said this for a long time about self driving cars. Like, everyone was really excited about them for a long time, and then they kind of didn't really work, like eight or so years ago. And there was this period where they were saying, oh, the models can't do green cones. We're going out there trying to take videos of green cones. And yeah, they can't do snow.
- [13:13]
C
So you're saying eventually you can just patch your weight.
- [13:16]
E
I'm saying that it worked for them, and so it might be possible. But in the case of language models, there's something a little more interesting happening because we now have two ways to learn. If you guys are ready, we could get into something a little technical, please, which I think gives you some insight. So there's essentially two ways you can teach machines to learn from data. One is called supervised learning, where the computer will copy what you did, which is basically what we're talking about now. And the other is called reinforcement learning, where the computer just does something and then you give it a reward if it does something. Well, and so for a long time, like the original ChatGPT was mostly just trained with supervised learning. It would just copy the text from all of the Internet. And so the best it could ever do is emulate Reddit posts very well. And there was a tiny bit of reinforcement learning, but people didn't know how to do it right. And then you mentioned this O3 model, which is kind of in some ways like a big jump. It made the models much better at math, much better at certain things. And the way they did that is actually through reinforcement learning. So they found out a way to kind of like, let the model think for a while and then give it a reward when it, like, gets the answer at the end. It's kind of scary.
- [14:28]
D
Yeah. When you say give it a reward, is it like, take a cookie paying robot? So she's excited about that? No, genuinely, like, what is the reward? You just tell it it did a good job.
- [14:39]
E
You just give it like a higher number and that makes it happy.
- [14:42]
D
All right. I'd get a little bit worried when we're like giving it cupcakes or something. Like, here you go. Good job. Just going back to the intro, you know, we were talking about how it feels like a lot of the progress on AI models is a little bit more incremental. And I guess it's hard to tell whether that's just personal bias because now we're used to them and the sort of wow moment has passed. But what does it feel like to you in terms of improvements? Are we seeing the improvement cycle accelerate or decelerate at this point?
- [15:13]
E
I think it's kind of like the market where it's like always, it gets faster for a little while and then it feels like things have slowed down and the progress is never quite in the areas that you expect. As one example, people really thought this year was the year when the assistants would start being able to act like actual assistants. Like the Year of Agents, people actually coined that term, I think, like the year of Agents. And it really. It didn't happen for whatever reason. Maybe it will in the next three months. But the agents are still pretty bad, the ones that you can use. But they did get way better at competitive math. Like now they can do these world class proofs that they couldn't do before. So it's almost unpredictable which areas the AI will kind of conquer next. But it does feel like progress is continuing.
- [16:00]
C
Actually. What happened with agents? I've never had a successful agent experience. Even basic things like come up with a list of every past odd lots guest.
- [16:09]
E
Oh, yeah.
- [16:09]
C
And put it in a file or something like that. Which there's an RSS feed that exists for odd lots. This should be really straightforward. Go all around and then something will happen or it'll get lazy. It'll hear like, here's 15. And what is actually. And this is. Thought leaders love this stuff. They love talking about the agent. So what actually happened with agents? Maybe they'll get there. But what is the roadblock there?
- [16:33]
E
I don't think there's any conceptual roadblock. Like, there's no reason why you couldn't collect data for that and train them either in a supervised way or using reinforcement learning. And it just hasn't happened yet. So I think maybe behind the scenes, it turned out that the problem was harder than people thought. Like, getting data from all those scenarios is really hard. And there have been some stories from, like, people that I've heard of that found these little companies in San Francisco, and they build these tiny environments for the AI labs to do reinforcement learning on for agents. Like, for example, doing a calendar. They'll build like a little calendar app, but make it have rewards so you can do reinforcement learning.
- [17:09]
C
And.
- [17:09]
E
And they can just sell that for like, hundreds of thousands of dollars. So I think the progress is ongoing behind the scenes. Like, there's a whole ecosystem built around it. It just hasn't really manifested in the products that we use.
- [17:21]
D
I was going to ask how much of the difficulty is the actual development of the models, the thinking part, versus just getting them to plug in seamlessly with other applications?
- [17:32]
E
Yeah, I think the second thing is probably the biggest barrier in terms of time. Like, it just takes a really long time to figure out what data and collect it properly and actually train the models on that data. But at the same time, there are people like me who are trying to work on better conceptual frameworks for training the models. So to go back to the O3 example, doing reinforcement learning on ChatGPT, like, that seems to me like a huge breakthrough. We didn't know how to do that before. It unlocks all sorts of doors and ways to train the models. So even if maybe you don't think that model was that much better than the previous one, it seems like it will give us huge improvements in the future.
- [18:30]
F
This message is brought to you by Nuveen. How would you invest if you knew the future was watching at nuveen? This isn't a theoretical question. It's a perspective that comes from navigating 125 years of market cycles, using foresight to innovate and adapt to the changing needs of investors around the world. And remaining steadfast in the pursuit of lasting performance for clients, communities and the global economy. Nuveen Invest like the future is watching. Visit Nuveen.com future to learn more. Investing involves risk. Principal loss is possible in business.
- [19:08]
G
They say you can have better, cheaper or faster, but you only get to pick two. What if you could have all three at the same time? That's exactly what Cohere, Thomson Reuters and Specialized Bikes have since they upgraded to the next generation of the cloud. Oracle Cloud Infrastructure OCI is the blazing fast platform for your infrastructure, database, application development and AI needs where you can run any workload in a high availability, consistently high performance environment and spend less than you would with other clouds. How is it faster? OCI's block storage gives you more operations per second cheaper. OCI costs up to 50% less for computing, 70% less for storage and 80% less for networking better. In test after test, OCI customers report lower latency and and higher bandwidth versus other clouds. This is the cloud built for AI and all your biggest workloads right now with zero commitment. Try OCI for free. Head to oracle.com strategic that's oracle.com strategic.
- [20:13]
H
Mint is still $15 a month for premium wireless and if you haven't made the switch yet, here are 15 reasons why you should 1. It's $15 a month.
- [20:23]
C
2.
- [20:24]
H
Seriously, it's $15 a month. 3. No big contracts.
- [20:28]
C
4.
- [20:28]
H
I use it.
- [20:29]
E
5.
- [20:29]
H
My mom uses it. Are you, Are you playing me off? That's what's happening, right? Okay, give it a try@mintmobile.com Switch upfront.
- [20:37]
D
Payment of $45 for three month plan $15 per month equivalent required New customer offer first three months only, then full price plan options available, taxes and fees extra.
- [20:43]
C
See mintmobile.com so you mentioned at the intro that it's possible. Hopefully you'll get a close but you might end up in a situation which you go to work for some frontier AI lab and we never hear from you again. Or you just post cryptic tweets like oh you have no idea what's coming or oh it's going to be so over or whatever. It's very annoying.
- [21:02]
E
Like the Death Star.
- [21:03]
C
Yeah, yeah, it's very annoying the way they all tweet. It's possible. Talk to us about like why not work on an open source project? And this is of course when people talk about deep Seq and a lot of the Chinese models that the US competes with, a lot of those are open source. Presumably you could keep coming on odd lots over and over again, why? Like, what is even the case for, for the best and the brightest to work on closed sourced frontier models?
- [21:29]
E
Yeah, it's a really hard question. I've struggled with this in my own personal decision making. I was originally thinking, oh, I'd love to become a professor and mentor younger students and get a whole like, group of these ideas going and start working on similar related problems to the stuff I was talking about. And I still think that would be fun. But there's a big gap in terms of the things we can do at Cornell and the things that you can do at OpenAI. They just have crazy infrastructure for training models really easily and, and data, precision data and a ton of really good data. And so I think as that gap has widened, I've felt like a lot of what we're doing is like kind of devising these toy scenarios where we can study interesting things. But I feel a bit disconnected from the real progress of humanity. You know, like, if you really agree that this is like the biggest problem of our time, I don't want to say it's like the Manhattan Project, but like, it's more like trying to go to the moon in the 60s. The space race. It's kind of like a space race going on in these different private labs. You want to be a part of it. Like, there's crazy energy in that. It has huge implications for the future of society. So I think I am interested in participating in that. My big question is like, if you think that the reinforcement learning thing was the most recent big scientific breakthrough, like 01 and then 03, what's next? And then like, where will that actually be happening? That's kind of what I'm thinking about right now.
- [22:57]
D
Just on the data point, I was reading your excellent substack and you argued that there's probably an upper bound to what you can get out of a given data set. And at some point, like the training starts to look pretty similar. Right. And the data becomes the differentiating factor. How important are data sets to AI research? And I guess, how do you go about finding really cool ones and what's left? Because I feel like using the space race analogy, everyone has been running so fast on this. It feels like all the data sets must have been explored by now, but I guess they haven't.
- [23:36]
E
Yeah, yeah. I think this is really getting to the heart of what people are trying to figure out right now in all these different labs. So I think you're pretty much right that all of the public data sets we have are pretty much used to train O3 or GPT5 or whatever. If there is a really good website that should have been scraped and downloaded into the model, it should probably be used. But there apparently is a much larger amount of private data than public data. I mean, you all work for Bloomberg, so you're probably intimately familiar with this. But if you think about the different AI labs that exist, they actually now do kind of have different data related modes like xai. They have all of the Twitter data that's basically impossible to get elsewhere. ChatGPT now has all of the user conversations with ChatGPT, which are really useful. Claude has a ton of coding data that other people don't have. Google has YouTube, which some people think might be the next source of making really good models. And none of those things are really included, at least not much in today's models.
- [24:40]
C
This is really important. Once a lab builds some sort of base, whether it's anthropic encoding or maybe cursor encoding, even though they're not like a core lab, et cetera, they become a source of their own data that literally nobody else has.
- [24:57]
E
Yeah, actually cursor is a great example. So they are very technical. They have really smart people. They're very small, so they haven't quite scaled to, at least in terms of the number of people.
- [25:07]
C
But I think about this, like every time when I've played with it, it's like, this is good, this is bad. I'm constantly teaching their model to get.
- [25:14]
E
Right, right, right, right. They're in a problem where they have the data. They just have to take the right algorithms and scale it up to train the model to be as good as Claude is. But that actually seems a lot more feasible than other companies that have no data and want to train good models, even if they know how. It seems very difficult.
- [25:31]
D
How closely are AI researchers working or talking to, I guess, other parts of the AI ecosystem? So, you know, chip makers, maybe cloud providers, that sort of thing. Is there a lot of dialogue or not really.
- [25:45]
E
I think certain people talk all the time to the chip makers. Like there's a big community of people. You know, the AI models all run on GPUs, and there are a lot of people that are getting really good at writing fast GPU code. It's called kernels. And all those people who work on kernels talk to the chip makers all the time. Like Amazon's making their own chips, Google has their own chip. Now all the hyperscalers are making chips. And I think they're all trying to talk to the People that actually write the fast code that runs on chips to figure out, I think they call it hardware, software, co design. Like everyone's kind of getting together and trying to figure out what the best way is to design the next round of GPUs.
- [26:23]
C
So you mentioned, okay, Google might have an advantage because it owns YouTube and there's just tons of, obviously just tons of data in there. So one way you could get access to, to the YouTube data is to literally be Google and own it. But another way that maybe you could get access to YouTube data is operate in China where there are no laws about this type of thing, or not be holding the US copyright and just sort of scrape it all again. Since most of the Chinese AI labs are open source, why isn't this just a huge advantage for the Chinese labs that they're really not going to be. Hey, OpenAI, they get sued by the New York Times. All these. Deepseek isn't having to deal with all these headaches.
- [27:07]
E
Yeah, I think the American AI labs will probably do things behind the scenes that they wouldn't tell you about to get good data.
- [27:17]
C
That's one solution.
- [27:18]
E
Just don't. Yeah, like I think they wouldn't release the models that are potentially trained on scraped or copyrighted data. But if that's the way to get better Math Olympiad scores, then people will, I think, I would guess, do that. But you're right that the Chinese model makers can just sort of take all the books that they can pirate from the Internet and train on them and they're not violating any laws and they can release the model to the public and it's all fine. Which is honestly great for us because then people like me could probably download a model that's better than we would get otherwise.
- [27:50]
D
What was your impression of deepseek when it came out?
- [27:53]
E
And now I was pretty surprised at how much of a splash they made. The model is really good and I think a lot of people are building on it, including me. And like most people that are at AI companies that aren't super, super big are building on Deep Seq. But it was surprising, like what a huge deal it was to people. Like, my mom's asking me about Deep Seq. I think my grandma knew about Deep Seek and she barely knew about ChatGPT.
- [28:21]
D
So that's when you know it's gone mainstream when your grandma start asking you.
- [28:25]
E
And there was nothing else so far, I think in the AI space that's made quite that much news.
- [28:30]
C
But it sounds like what you're saying is that It's a very good model, but that on the actual specs, from your perspective, it didn't quite deserve as much attention. Like it was good but like in your view it's not so good that everyone needed to be talking about it at the time.
- [28:46]
E
Yeah, I think it's really useful because they released all their model weights and they said exactly what they did to train it, although they didn't say what the data was. But it gave me the impression they're maybe 6 to 12 months behind the American AI labs in terms of how well they can do the training and stuff. But it still was a pretty big update for me to know that, wow, there are 100 people that don't have PhDs working at a Chinese hedge fund that are training these like cutting edge models. Like it is incredible and they work very hard. They're very good.
- [29:32]
B
Join Bloomberg in Houston or via livestream on November 4th for the Future Investor Finding the opportunities this 2025 event series will examine how companies are investing in their businesses to create efficiencies, innovating their products and services and improving the customer experience. This series is presented by Invesco. Q. Q. Q. Register@BloombergLive.com FutureInvestorHouston that's BloombergLive.com FutureInvestorHouston do you.
- [30:02]
D
Have pressure or do you feel pressure? Do AI researchers in general feel pressure to consider monetization when they're researching things? Or is it, you know, mostly still curiosity driven, that sort of old school Silicon Valley, we're improving the world kind of thing? Or is it much more mercenary given that all these big companies seem to be competing in the same space?
- [30:26]
E
Yeah, I think that over time it's gotten harder and harder to do things that are just like cool ideas or seem cute but don't have any necessary application and things are getting closer and closer to products, you know, even like the language models that power ChatGPT. I was working in those before ChatGPT and they had some uses but also they're intellectually interesting and like fun to build. But now if I came up with a better way to train ChatGPT, that's like a multi billion dollar innovation.
- [30:59]
D
The stakes are higher.
- [31:00]
E
Yeah, I'd be like an asset to like the United States government or something if I knew how to do that. So I guess it depends on what kind of problems you work on. Like I'm more interested in understanding how things work so it becomes a bit less financially dire.
- [31:14]
C
I think that 6 to 12 month gap between what was that that was a January deep seat moment.
- [31:21]
D
Yeah.
- [31:21]
C
Although really like everyone should have, it was in December that they first got attention, then for some reason really hit in January. Is that a sustainable gap? Is there something either in access to data, access to talent, access to compute, access to chips, whatever, access to energy that in your view will allow US frontier labs to maintain some sort of 6 to 12 month gap for a while?
- [31:43]
E
It's pretty unclear to me. I think there are different beliefs you can have. You can believe that the ideas and the people are really the thing that differentiates the models. And in that case, I think we haven't so far seen a lot of like the top US AI researchers going to work at Chinese labs. So that seems stable. You could think that chips really matter. And in that case the chip race is really happening between big American companies. I think actually China has a pretty big deficit coming up in terms of the GPUs we're exporting or you can think that the data matters. And I guess actually any of these point in the favor of the U.S. i think if you think the data really matters, maybe the data they gather through deepseek.com usage is really good and they can use it to bootstrap a better model. But I think the American company companies really do have an advantage. Like you all might have heard this story just as an anecdote. Apparently at Anthropic they've been buying and scanning thousands of old books for several years. So they have this division, I think they're based in New York, that buys like shipping containers full of old manuscripts, cuts off the spines and puts them in these scanning machines and then they turn them into like really high quality text. And so I'm noting Claude has this like weird aspect to it. Maybe part of the reason is they've gathered like trillions of words worth of old book data over many years and that's pretty hard to replicate elsewhere. So I think that head start really does mean a lot.
- [33:10]
D
What are you most excited about at the moment? The book thing sounds very cool, but what is getting all your attention right now?
- [33:18]
E
Thanks for asking. I think I mentioned before, I'm really trying to figure out what's coming next. There are some obvious things like we can get computer usage data and train better agents, or we can get more coding data and make them better at coding or writing GPU code or whatever. But what are the non obvious advancements? And my personal opinion is that the next round of improvements in AI models will come from some type of personalization and online Learning, which means models that one are trained per person or per company. So you could think of. ChatGPT is the same model that gets served to everyone. So it has to store information about random restaurants in countries you never go to. But instead, if you had a chatgpt that's specific to Bloomberg or specific to your work, it might be able to use more of its brain to do work for you. And then the second thing is if it was updated every day. So if you ask it to make your odd lots calendar or RSS feed and you're like, no, that was wrong, you did it wrong for this reason, this reason. And you try again tomorrow, it'll still break tomorrow because it doesn't continuously improve its capabilities. I think that's the direction things are going.
- [34:36]
C
I've heard people talk about this. Now, granted, models are getting better over time, but people might compare a coding model to a beginning software engineer and say the coding model is better, but that software engineer is going to start getting better the next day they're on the job, and every day for the rest of their career, they're probably going to be a better software engineer than they were the day before. Whereas at least that version of the model will not be better. Is that right?
- [35:02]
E
Yeah, yeah, that's right.
- [35:03]
C
That seems like an issue that people talk about in your world.
- [35:05]
E
Yeah, yeah. I think this is a big problem. It's like we have to wait six months for the ChatGPT 5.1 to come out, and then maybe they'll include your problems as the training data. And so maybe it'll get better, but it might not. And instead I think people need to think about ways to do that update more dynamically, like every time you talk to it, or maybe every night when you go to sleep, the model kind of gets to work and studies what it was talking to you about and crafts better tests for itself and then learns. And then when you wake up, the model is actually better.
- [35:35]
C
The other big question that I have, and it's kind of related to this, especially when we're talking about AI replacing humans and certain forms of labor, is that, like, do we need really, really advanced AI? Like, in other words, like, there is a lot of, again, the existing models are extremely impressive. Like, in your view, do we need to get a lot better technically for them to have economic impact? And since these are in many cases businesses, at the end of the day, is it necessary that there's so much work being done towards advancing the cutting edge?
- [36:11]
E
Yeah, yeah, that's a great question. Like, we could have really good interns without ever getting better scores on the Math Olympiad. Like, that's not necessarily something that we ever had to go after. I think part of the reason for that is the AI labs are engaged in this kind of neck and neck race to have the smartest model. But I totally agree that AI could be economically transformative without having a higher ceiling in terms of what it can do. It's more like it needs to be more consistent or like dependable than actually smarter.
- [36:41]
D
This might be a weird question, but once you've made a sort of foundational improvement to a particular model, how easy or difficult is it to rewind if you need to? And one of the reasons I ask is because, you know, some people have been complaining that they've been training ChatGPT to, I don't know, be their boyfriend or whatever, be their therapist.
- [37:02]
C
This is a legit topic in my opinion.
- [37:04]
D
And then it gets upgraded and all of that training suddenly disappears and the personality of the model changes.
- [37:11]
E
Yeah, that was a really interesting story. So I think the model before GPT5 was 4.0. And they said that they thought internally, like all the scientists and coder people, that the new model was superior in every way. It gives you shorter responses, it's a bit nicer, it's much smarter. And then people got really upset because they had spent so much time talking to the old model that they felt like they'd experienced like a serious loss in their life.
- [37:37]
D
Joe would miss the. Love it.
- [37:39]
C
Yeah, love it, love it. No, but for real, this is unironically. This strikes me as another example for open source, which is that I. If I'm going to form a. I don't see, I'm 45, I'm too old for that. But if someone is going to form like some sort of friendship with an AI model, I don't want it to be at the whim of Sam Altman deciding like, oh, there's an upgrade. I would like to be friends. It's also weird to be friends with the model that I know that I can run in perpetuity and it will never change.
- [38:11]
E
Yeah, I think that's definitely a good argument for why open source is important. And if you ever fall in love with a model, you should fall in love with an open source model.
- [38:20]
D
That's good life advice, practical life advice.
- [38:22]
C
That is really good life advice.
- [38:24]
D
Well, speaking of open source, you know, I know programmers tend to like open source for obvious reasons, but are there any downsides to open source for AI specifically?
- [38:35]
E
I think if you're running a company. There are a lot of downsides potentially to open source. If you have some brand new fancy way of doing computation inside the model that's actually better, you might want to keep that information to yourself. And when you release the model to make it runnable, you have to release all the code to run the model, which might contain your secrets. And so I think that's why people are hesitant to do it. The other reason is because when you release the model, it actually contains quite a lot of residual information about how you actually trained it. Like you might be able to infer what the data set was and what the training process was, or even reconstruct the entire training data set given just the weights of the model. And so if you're worried about people finding out that a certain thing was in your training data, you probably can't release that model open source.
- [39:24]
D
That reminds me how much of an AI researcher's day to day life is just like looking at other model, other people's models, and trying to like, I guess, pull them apart and figure out how they were made and sort of work backwards.
- [39:39]
E
That definitely happens from time to time. I think usually the scientific process is something like you start with other people's models and you run them and you see what happens and then you decide on some part of that process that you think could be improved or could be explored further, and you make some tiny changes to it and then you run it again and you compare numbers or you make graphs of what happened before and what happens after. So actually quite a bit of it. Like for example, the GPT2 model from OpenAI which was 2019 or something, their first kind of really larger scale chatbot. I've spent hundreds of hours kind of playing with that code and talking to the model and stuff like that. So thank goodness for open source for that reason.
- [40:21]
C
I joked in the beginning about you having $100 million salary, but for real, as you think about your career, and I hope you do get a hundred million dollar salary, but as you think about your career, what excites you and how much is it money? The reason I think about this is like there are huge checks out there, but maybe some things are more, maybe achieving AGI is more excited than making an ad network more efficient. Maybe something, there's something more exciting than shaving off a billionth of a second in terms of a trade execution. All these things like how much is it about exploring the frontiers of science, the new space race, landing on the moon versus the paycheck.
- [41:04]
E
It's all about the paycheck. I'm just kidding. No, no, not at all. Yeah, it's funny you ask. So this hasn't happened to me, but just in the past two weeks or so. A good friend of mine has been dealing with this problem because she got an offer on the order of like tens of millions of dollars per year from a big AI company and she wasn't sure if she wanted to work there. And I think originally she said no and then they doubled her offer and then like it's the exact same amount of cash but twice as much per year for a certain number of years. And you know, we were talking amongst ourselves, like, what does this even mean at this point? Like you're a 28 year old computer scientist that's been coming from a PhD, so you make more on the order of tens of thousands of dollars per year. I honestly think personally the marginal difference between having like 10 and 20 million dollars is like very low. Like I don't even know what I.
- [41:56]
C
Would do with this is. I, this is my experience. For me, making 10 million, 20 million is basically been the same to me.
- [42:03]
E
Yeah, congratulations. But so yeah, I think like there's more of a desire to like be there the next time something really interesting happens that kind of supersedes the money. Like any of these places will pay you what's like a really good salary to live on. And so it's actually not a big consideration. It only becomes complicated when you have like one option that's going to pay you like 40 times more than the other option and then things get confusing.
- [42:30]
C
No, this is, this would actually Joe's.
- [42:33]
D
Just thinking about making 20 million.
- [42:34]
C
No, I think like, because I think about, okay, what if you have this great salary and you're like, can live very easily in New York City and have a really great life. Or you can make 10 times that, which is a stupid, insane salary.
- [42:48]
E
Right.
- [42:49]
C
But you don't really like your job. But it's so much money that strikes me as like not a trivial life. You only live one time. There's like a diff. So it could be a difficult question.
- [43:00]
E
Yeah, yeah. But you can remind yourself that like the job you take once isn't the job that defines you forever. Maybe the right thing to do is to take it for a few years, but not the whole time and then go do something else.
- [43:11]
C
Everyone says they're going to do that.
- [43:13]
D
And then they get locked in.
- [43:15]
C
Yeah.
- [43:15]
D
Speaking of insanely large salaries, we know that people are earning these salaries because they're like star AI researchers. How much does personality play into where you want to go work? Would you want to go work somewhere specifically because there's an absolutely amazing researcher, or does it tend to be, again, more about the paycheck, maybe more about the data that's available to you, or maybe more about the specific project that you're going to be working on?
- [43:43]
E
Yeah, I think different people assign different amounts of weight to each of those things. In my experience, most of the people I know come from academia, which means they already kind of gave up more of a salary to do study things more deeply for several years. So I think people that I know are more biased against money, but people do care about that. But I think that the ego thing really matters. And so some people want to feel like they're very important and they're working on a problem that matters. One way some companies are able to pull researchers away from other companies is by saying, we'll assign you more importance in your role and we'll give you.
- [44:15]
D
We'll give you a really big title.
- [44:17]
E
Yeah, exactly. Seriously, the title is like, okay, maybe before you were, like, a researcher. Now you get to be, like a head researcher. You get to have people under you. You're a chief scientist, and all these things do matter to people.
- [44:28]
C
It's a very good book about pursuing a mission in the realm of a driven visionary, even when it's commercially.
- [44:37]
D
Just say it, Joe. Just say it.
- [44:39]
C
Yeah, that's right. No, I think about this all the time. It's like, do you want to work for Ilya or do you want to work for Sam? And which one is the Ahab? And which one is just trying to make an honest living selling ads? I find this to be, like, a genuinely interesting. Interesting question for any individual to have to reckon with in this career.
- [44:55]
E
Oh, absolutely. And sometimes it can be very difficult to tell.
- [44:58]
C
Jack Morris, thank you so much for coming on. On. Please pursue a career that will allow you to come back on odd lots.
- [45:04]
D
Or insert the odd lots clause when you're negotiating your $100 million salary.
- [45:09]
C
Yeah. Or take the 50. Say, you know what? I'm 50 million, but let me. I don't need the 100 million. 50 million, but keep the Al.
- [45:16]
E
Yeah, that would be fine with me.
- [45:17]
C
All right, great. Well, thank you so much.
- [45:19]
E
Yeah, thanks.
- [45:20]
D
Thank you so much. That was great.
- [45:34]
C
Tracy. I think about that sometimes. Like, what if you got, like, an insane salary like that? You just couldn't. You'd be insane to say no to. But, like, I don't know. That's I mean, it's not our problem, but like, wouldn't that be fun? You know, it's like, oh, but you're going to be working on ad optimization or whatever, and you're not going to be there when they land on the moon. But you got paid 10 times more than the people at the base station working on landing on the moon. That strikes me as kind of a tough life choice.
- [46:01]
D
I think you're using up a lot of brain power and energy on a problem which, as you said, is not yours.
- [46:06]
C
That's exactly right.
- [46:07]
D
No, that conversation was really fun. Nice to talk to an actual researcher who's doing stuff in the space. One thing I thought was very interesting was this idea that everyone gets excited about a specific improvement in AI and then it seems like that particular one doesn't materialize and instead something else emerges as, like, the big breakthrough. So instead of agents, we have math.
- [46:31]
C
And math, which none of us will ever need. I would really like for an agent to do something simple. I'm going to a city, book my new trip or whatever, or change my flight. Oh, my God. I tried to change my plan.
- [46:41]
D
Oh, that would be amazing recently.
- [46:43]
C
Change my flight. Here's my information. I would like that. I do not need the Math Olympiad. I am very impressed. I don't need it.
- [46:52]
D
Also, I am now very, very intrigued by reinforced learning and how you actually reward the computers for doing good stuff. I feel like actually that would be a really interesting area to mine, which is motivating the models to do better.
- [47:08]
C
Yeah, I've thought about that. Like in chess, like, how do the computers know they want to win? Why do they care all these things anyway?
- [47:16]
D
Why are they here? Why are we here?
- [47:18]
C
That's the thing with AI conversations gets existential fast. One thing we didn't talk about, which I am interested. No one really talks about AI safety anymore. Have you noticed? Like, like very little. Like, for better or worse, you don't hear people, it's just all money. And they don't really talk about, will they kill us all one day. But one thing I did wonder about, so when Deep Seat came out, one of its breakthroughs was it showed the whole chain of thought, right? You could see that. Which prior to that open air chatgpt chain of thought model, didn't show you that.
- [47:46]
D
Right.
- [47:47]
C
And it does strike me that if there are certain things that are for safety reasons or whatever held back or they don't want to do this, the nature of competition means all the guardrails are coming off eventually. Like that's if there's some guardrail you have on, someone's going to open source whatever it is and they're going to all give it up.
- [48:05]
D
Yeah, both on the guardrails and on the data usage points as well. All right, well, shall we leave it there?
- [48:10]
C
Let's leave it there.
- [48:11]
D
This has been another episode of the Odd Thoughts podcast. I'm Tracy Alloway. You can follow me at Tracy Alloway.
- [48:17]
C
And I'm Jill Wiesenthal. You can follow me at the Stalwart. Follow our guest Jack Morris, he's at jxmnop. Follow our producers Kerman Rodriguez at Kerman, Armand dashiell Bennett at Dashbot and Kale Brooks at Kalebrooks. For more Odd Lots content, go to bloomberg.com odd lots with the daily newsletter and all of our episodes and you can chat about all of these topics 24. 7 in our Discord, Discord, GG Oddlauts.
- [48:41]
D
And if you enjoy Odd Lots, if you like it when we talk about $20 million salaries that will never be ours, then please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, you can listen to all of our episodes absolutely ad free. All you need to do is find the Bloomberg Channel on Apple Podcasts and follow the instructions there. Thanks for listening.
- [49:34]
I
Check out the new season of Reasonable Doubt, now Streaming on Hulu, L.A. s most successful attorney Jack Stewart, defends a young actor accused of murder. Follow Emma Yahtzee, Coronaldi, Morris Chestnut, Joseph Sakura and guest stars Kash Doll and Lori Harvey as they fight their personal battles in the spotlight of the year's most sensational trial. In the pursuit of justice, every move counts. Reasonable Doubt Season 3, now streaming on Hulu and Hulu on Disney for bundle subscribers. Terms apply.
- [50:04]
J
In the heat of battle, your squad relies on you. Don't let them down. Unlock elite gaming tech@lenovo.com Dominate every match with next level skill, speed, seamless streaming and performance that won't quit and push your gameplay beyond limits with Intel Core Ultra processors. That's the power of Lenovo with Intel inside. Maximize your edge by shopping@lenovo.com during their back to school sale. That's Lenovo.com Lenovo Lenovo this Prime Big Deal Days October 7th and 8th. You can get a great deal on a new foot spa, transforming you into the queen of kickin it.
- [50:42]
F
Wait, this has bubble jets.
- [50:45]
J
Okay, get up to 40% off with prime big deal days October 7th and 8th.