Where Are AI's Productivity Gains? — With Ethan Mollick

0:00If AI is such powerful technology, where are all the productivity gains? Let's talk about it with Ethan Mollick, the author of Co-Intelligence and a Wharton professor that we quote all the time on this show. That's coming up right after this. Welcome to Big Technology Podcast, a show for cool-headed and nuanced conversation of the tech world and beyond. We have a great show for you today. We're joined today by Wharton Professor Ethan Mollick, the author of One Useful Thing on Substack and his great new book Co-Intelligence, which [music] I highly recommend. Um and today we're going to talk all about how AI's uh um improvement on the technology side can

0:37actually lead to productivity gains across [music] the economy. We'll also go through whether the technology can keep progressing, whether it will hit a wall, and a number of other topics. I'm thrilled to have Ethan here because [music] uh we we quote you all the time on the show. I've been reading you for a long time. You've been on the show [music] before. Uh your book is great and we have so much to talk about. So, thank

0:56you for being here. Thanks for having me. Um All right, let's start with AI and productivity. You know, AI is a very powerful technology. We've all seen it if we've tried it. Um the question is what that power leads to. And the the um conventional wisdom is if AI is actually that useful, we should see the

1:17impact in productivity stats. We're not seeing it. Um this is from the Dallas Fed. US labor productivity has grown at an annualized rate of 2.4% since the beginning of 2004. Not a huge percent. It's not revolutionary. Like people like Dario have said we might see 10% increases uh in GDP. We haven't seen that. And um the most astonishing thing to me in this Fed study is that they um charted out high AI exposure and low AI exposure jobs. And now high AI exposure jobs are seeing more productivity gains than the low AI exposure jobs, but they're actually seeing less productivity gains than they saw during COVID and even less productivity gains,

1:59than they saw pre-COVID in at some times. Um, the spikes just do not reach the same heights. What's going on, Ethan? I mean, I I think anyone who's in an organization probably has the right answer to that problem, which is you don't get organizational productivity gains instantaneously from individual productivity gains. Um, you can be a hundred times produce a hundred times more code than you did before, and that's great if you're a solo entrepreneur producing a lot of code. If you're in an organization that's doing two-week-long sprints and has to coordinate with marketing, and has stand-up meetings every day, which is a rational way to to to organize work,

2:31that hundred times productivity gain is actually can be a barrier, right? No one has time to deal with all of that code. And we see that, too, in shipping stats. Two hundred you know, twenty times more code being produced, uh, thirty percent more products being shipped, right? And so, it's not an uncommon problem. We

2:46haven't adapted organizations to AI yet. Yeah, I mean, you there's there've been some stats, I think you're referencing this, um, in the App Store, where there've been many more apps shipped than previously, but uh, even fewer reviews than before, right? So, it's like, um, the productivity, quote-unquote, or the the, um, work looks like there's been more done, uh, but when it actually comes to like usage or value,

3:13it's so far not showing up. Well, well, actually, those are two different things. So, one thing is the flood of stuff, right? And we're seeing that everywhere. So, there's a lot more novels, a lot more games, a lot more apps. But, I'm actually talking directly, there's a nice set of studies looking at product shipped, right? And we're seeing huge increase in the amount of coding, smaller increase in the number of product shipped. There's still an increase, but it's less, right? By an order of magnitude. And the reason for that is, again, we don't have multiplayer AI. AI is a very single-player experience right now, where you get individual performance,

3:42but organizations are about many people performing at once. Yeah, I've I've thought, when it comes to these capabilities, that the companies that are really going to, at least in the near term, pull ahead, are going to be the ones that say individuals who are using this technology, um we're going to empower you to actually go ahead and, you know, break free from some of the um standard red tape that would hold you back in a typical, um you know, corporation. Like, um I think in most companies if somebody uh

4:10is It's It's interesting what happens. Someone comes in after they're hired, they're they're given a mandate, they tend to get beaten down if they have a lot of new ideas, if they want to change things. At a certain point the companies are like, "No, no. You're actually not here to have ideas. You're here to sort of play this role." And so, I think we might end up seeing a divergence in terms of uh competitiveness among companies that enable people to actually run with their ideas and then companies that follow this traditional blueprint of like, "Oh, you have ideas? You want to get things

4:39done? Just, you know, do do your job." Right. I mean, in in the book I talk about leadership lab and crowd being three things you need aligned. You need a leadership team that actually buys into this and starts making organizational change. You need a crowd. You need a people who are using and inventing ideas and that's often the line employees are talking about. And then you need a lab. You need to actually be doing research on how to take this stuff and and spread it out among everyone. I think the only thing I'd push back on a little bit there is

5:04we still have the multiplayer problem. Like, if you're still if you're a hundred times more productive coder, we still need to figure out how do we give you the roadmap to follow in. We still need the marketers to work with you to figure out how you get the material shipped out. We still need to figure out how to maintain the software. So, it's a little tough to just say if, you know, here's productivity gains alone from individual players even if they're given the freedom to explore, we still need to

5:24turn that into process. Yeah, but this is not year one after ChatGPT. Like, if we were talking in, you know, October or November uh 2023, I would say, "Yeah, that makes sense." Uh companies have had four years to adjust to this. Yeah, and the capabilities are getting better, but they've had a long time to adjust to this. And there is an urgency for them to figure it out soon because the magnitude of the bets on uh building out the infrastructure and the products and the investments in the companies, you know, that money eventually has to be returned the investment has to be returned to a degree and

6:02if you don't have productivity gains by definition it seems like there you're not going to get the return on the money. Well, I mean we're talking about top-level productivity gains in the US economy. I talked to lots of CEOs of very large companies. There is almost nobody at this point and the data from my colleagues back this up who isn't seeing some ROI from AI. Like it's not this is not an issue anymore that people aren't seeing ROI. What they are trying to do is figure out how to move from individual productivity gains and you know, $20 here, $30 here which sounds like a lot to us but it's not a lot to a

6:29large company to how do we transform our organization around this, right? So I don't think it's a question anymore of are you getting gains? I don't hear that from C-level people. I did a year and a half ago or a year ago. That's no longer the question. The question is what do we do now, right? So we have a you know, we have 100,000 employees. Like I think it organizational change doesn't happen that quickly. I mean ChatGPT came out in 2022, right? GPT 3.5 and like companies just started

6:56adopting this in the last couple years. There's no playbook, right? Nobody knows what to do in this case. All the traditional methods of like the consultancies that would help you, the enterprise software companies that would build this in. I mean it is a revolutionary change. So I think that organizations are absolutely seeing the ROI. I'm not I haven't heard of a single organization saying that AI isn't paying off at a large scale right now but they're not getting the transformation yet but that requires management to make decisions about what's going on and they've barely begun to do that. Many of them aren't even still using advanced

7:24systems. But okay, they are while while some might say they're getting this ROI, they're also telling their employees to pull back on the spend and this era of you know, we're going to spend tokens freely, you know, that's that's ending. And so yes, there's some ROI, but that what their actions suggest is there's limited ROI right now. I'm going to I will disagree with you again on this. Like, there is So, first of all, there was a we keep whipsawing between different views, right? Like, token maxing was big this spring, and then it was like, "Oh, no, everyone's spending too much on tokens." It is becoming a

7:59fast-growing expense. Companies are worried about this. Most When I talk to a lot of Fortune 50 companies, though, they are viewing this as something that is an R&D expense to try and get under control, to figure out what's going on. A lot of them don't know in the same organizational change ever Like, I was talking to a very large company the other day, and they're saying we have some people who are 100 times more productive, and some people aren't really getting any benefit out of AI. We don't know yet how to spread the 100

8:22times more productivity to more people. So, I I think that there's news articles about people pulling out. Now, most of the people who are talking very publicly about how they're cutting AI spend tend to be people with competing AI products. So, like, Microsoft and that Microsoft, Meta, and Palantir are telling you, "We're spending too much money on Claude." I don't know if that tells you that much about the world, right? Our revenues are still growing for these companies. Absolutely, nobody has cost under control, use under control, but there I don't see any sign of people saying they're pulling back

8:49because AI turned out to be useless. Not useless. It's just the question is how much return are you getting on the investment? companies are getting large returns that we're talking to, but they're also in experimentation and evolution phase, right? It is You can't get returns without making changes. Most organizations I know know that. Making changes takes time, right? And they need to figure out how to swallow this new technology. So, I think I think that it's there's too much of a sort of back and forth like, AI's changing everything, AI's changing nothing. When

9:16I look inside organizations, right? C-level people are deeply, you know, spending a lot of money and a lot of time trying to figure out how to make things operate well. I'm seeing transformation in lots of organizations, including really old-school biopharma companies and financial service companies, very large ones, who don't want to talk about exactly what they're doing yet cuz it's a source of competitive advantage. And I think people are worried about token costs,

9:36but that's often a management problem. Like if you do if you don't make any decisions about how tokens should be used, right? Who should be using them? What process should be using them? Then and you would incentivize people to use as many tokens as possible, they'll use as many tokens as possible. That it feels like like I I feel like we're kind of in this chicken and egg where we're like, you know, where is where is the return? It's I keep telling you it's there um because companies keep telling me it's there and surveys show 85% of companies have positive ROI from AI um as of last year. So, I think though the

10:05major transformation hasn't happened yet, except it's starting to in some organizations. But again, you have 100,000 employees, how are you supposed to even think about mapping out all the processes in your company? How do you start deciding about how who gets to use tokens under what circumstances? How do we start to adopt the gen-tech models when they're basically less than a year old? There's a lot of stuff happening

10:23all at once. Yeah, okay. So, let's go bit by bit here. So, first of all, you've said that the companies are already starting to see ROI. Um where are they seeing that ROI? Where are the areas that they're seeing it? So, in terms of so, productivity and cost savings are already there. I've actually been worried that companies are pushing too much on productivity as the main goal. So, my colleagues at Workday have been surveying 1,000 large companies and like I said, last November, a year ago, um there 85% of large companies with over a billion dollars in revenue reported positive ROI from their AI use, right? We don't have detailed breakdowns

10:56on every piece of whether it's cost savings or spend, right? But you must have some anecdotes of like a company put you know, AI into action and Yeah, I mean sure, but but Tell us a tell us a tell us a story. so a lot of them I can't even talk about necessarily. I mean, so I talked to a you know, there's a another Fortune 100 company that had they gave its senior leadership team access to basically open clause with access to the full company data and

11:21completely changed how management works. Like C-level teams are using this as a super analyst to find problems inside the company they weren't able to identify. I have spoken to a large financial services company that has launched a whole bunch of new product lines that they never had before as a result of AI. Talked to a couple a couple of very large like hedge fund and PE funds that are seeing huge returns to management technology right now and all the companies are growing faster inside the portfolio. Like lots and lots of cases like this, right? Often times they don't want to talk about it in huge detail and

11:54they're all still in the experimental phase. I mean a lot of these are outside of pilots now they are rolling out at the organizational [clears throat] level. But it's it's tough because it's a complicated kind of you know set of choices at the company level rather than a product you adopt and get an instant

12:07return from. You know a stat that we used to quote often on the show was that it was like one out of every eight or one out of every 20 something in large a large number of AI pilots were out there but only one out of every you know eight to 20 or so actually made it into production. Have you seen the numbers sort of shrink in terms of the pilots

12:26that it takes to get to production? So two things that is a famously bad study. Let's hear it. Let's hear it. So so that's the MIT Nanda study from last year and at the time it was it's somebody went out and just interviewed people at a conference and coded that himself. No that was that was 85% of companies were like

12:45No that was 95% of pilots fail. Yes 95% that was the famous sort of Nanda study. You've actually been vocal about how the study hasn't been good but there have been other reports that have come out that said you know it's not that maybe that was the one in 20 but that the more reasonable reports say

12:58like one in eight or something. I would hope that that's the case that most AI pilots fail. Again the one thing I'll keep emphasizing again is I I you know I don't take money from the AI labs. I don't really care whether they succeed or fail individually. The you know the transformation that's happening in companies is pretty inevitable when I talk to organizations at this point and I think they have to be doing experiments that fail because if they don't do that how are they going to

13:21learn this? There's no playbook. So I'm hoping for high failure rates otherwise you're not engaging in experimentation. Yeah, all right. Let's Let's now go to the next step, which is what happens What has to happen for us to see that widespread change within companies where we get those big productivity gains. So, it's a two-parter for you. One is how it you know, the type of change that unlocks

13:45the full potential. What does it look like? And what do we end up seeing if companies make the moves that you suggest? So, I there's no one move, right? It's about thinking about structural change, which takes years for for a lot of organizations to do. And so, what I'm seeing So, examples of things that I'm seeing on a sort of structural level, very common situation is pulling apart organization teams that existed. So, called disbursement, you pull out your best senior engineer, put them with subject matter expert and a marketing and strategy person, and then have them do like impossible projects. Like I just spoke to a company, they went from a

14:20$200 million data migration plan for their old, you know, old mainframe data to they renegotiate to a $20 million plan just a few months ago, and the CTO is thinking like, well, maybe I can do this for $2 million in a couple months. Like real noticeable change, but it involves literally pulling apart teams and thinking about what do we want to do and resetting projects from the

14:40beginning. So, how does a team How does a team typically change? So, because because individuals are now starting One of the other things we're seeing in companies is a dissolving of company boundaries, right? So, coders are doing design work, designers are doing coding work, product managers are doing everything. So, part of what's going on is how do you reassign roles that way? Some companies are assigning everyone the role of builder, right? And that's that's what they're doing with it

15:04with AI. Um but other organizations I see are trying to think about how do we take teams of highly talented people, these get people already getting returns from AI, put them together, give them really hard challenges to do, high token budgets, and seeing how far they get. And if they fail, which happens some of the time, then how do we learn something from this formally rather than just burning tokens and

15:23setting them off into the air? Is that so different from the way that companies function today? Yes. I mean, most organizations do not have uh are not doing far exploration work. Mhm. Um and especially in organizational design and form. So, the 40% of the of the advantage of American companies in 20th century came from better management according to paper by Professor Sudhir and other people from Harvard. Um and we don't we used to do a lot of exploration of organizational form, right? Companies were experimenting all the time with how to organize themselves, how to get work done. I mean, a large part of the story of the 20th century was the story of

15:57American manufacturing and firm organization transforming. That is not happening as much last 30 years. People use outside vendors like, you know, that SAS vendors will tell you how to run part of your business, use consultants, and your R&D is focused on your customers or outcome. So, a lot of organizations have very little um muscle memory left of organizational change. So, it is actually pretty rare to see organizations play with fundamentally re-dividing how work operates, how roles operate, what uh mandate people have to get work done. I mean, that doesn't feel like normal

16:27operation of business to me. Okay, so I want to ask you and by the way, um so I know where you stand cuz obviously you wrote it down in a book. I'm just trying to sort of um talk through your arguments. So, for those who are like, you know, speaking about me, this guy, you know, I'm I'm just trying to No, it's fine. There's nothing wrong I I am I am

16:44appreciate that you're going for. No, I'm excited like I mean, and I I should point out we don't have all the answers, right? I mean, but also we don't have all the answers sounds like an academic result. Nobody knows what's going on is a better way to phrase this, Yes. Right? Like there like I I think the thing that I'm reacting to a little bit is this sort of I don't think skepticism about AI's ability to create transformation warranted at this point. I think that's a dead end. I think um anxiety about how we shape this in the right direction is warranted and I think how do companies start to think

17:16about this in a way that helps large companies start to transform? Is it very reasonable thing to worry about and why we haven't seen those results? I spent a bunch of time in the book talking about why haven't we seen transformation happen. So, I'm very much on your page and I'm not going to defend that trend like, "Oh, no, AI has changed everything today." Part of the interesting puzzle of the book and why we're in the phase of coexistence, right, and stuff like that is we are trying to negotiate that right now. What does the world look like so we get productivity gains and good

17:42results for society. Okay, great. So, then I'll keep pressing. So, thank you for the permission to do so. I mean, um I I get it if you're talking about industries like consulting, software, but I wonder I I don't even want to use the term real economy to to describe what I'm about to describe, but like there are parts of the economy where um shifting the organization in the way that you you talked about, adopting AI in the way that you talked about, um I I'm curious where you end up what you get at the end of the line, what the output is. I'm talking about like, let's

18:16say like Frito-Lay adopted AI like crazy. Like, do we just get like a new flavor of chips? Or or a car company adopted AI like crazy? Do we get like a new vehicle? Or a small hotel chain? Um do they build a new hotel? Like, when it comes to these type of businesses, which I think are the heartbeat of the economy

18:36to some degree. I mean, uh services are the heartbeat of the economy right now. Yeah. But but if you want to talk manufacturing, I mean, I literally talked to some of the companies in that exact range that you're talking to, right? And again, I can't name names not cuz I'm trying to be cagey, but because I don't know if we should talk about it or not. Um but I I you know, I mean, there's an endorsement in the back of my book from the CEO of the just stepped down CEO of Walmart to give you a sense of the kinds of companies I talk to, which I think most people would consider

18:59real economy stuff. Um and the the first of all, there is a huge amount of change in these organizations. Some of this, by the way, is R&D work, right? Like, literally I know the equivalent of the better kind of Frito-Lay chip, but that isn't an expensive process inside companies that make new products. And speeding up new product development is actually a big deal to them in a big way, right? And finding new trends. Um I've been talking to large-scale manufacturing and agribusinesses that are thinking about things like how do we use the fact that AI can now look at cameras to figure out new things about where which fields are

19:29in trouble and which ones aren't. We like we manage construction projects the same way we have since the 1940s. What does it mean when you have an AI that can give people advice about what to do differently and can actually give information about what's happening on the site or not. You have really sets of really interesting sets of changes happening everywhere, right? On the hotel side like or even in the Walmart store, they've been pretty open about this set of stuff. How do you give advice to everybody to a senior store manager who needs help figuring out what

19:54to do at the store when there's trouble? Like management is a real thing here and there's pretty transformational things. That's aside from AI for planning and AI for central management and AI for HR. So, I mean, I I actually think, you know, both services are the real economy, right? And it's a big deal of coding, you know, I think one out of every, you know, one out of every 16 jobs is is management in the US, you know, what and coding isn't follows soon afterwards. Like these are big jobs that represent a large of the economy, but also I think the whole point of this is

20:23general purpose technology with widespread effects. I mean, take even the most narrow thing, right? I was I was already talking about migrating databases over at large financial services companies and mid-size banks that have COBOL still running internally. These have literally been the projects that have been the backbone of the of a large part of, you know, IT for a long time, those are being transformed overnight. So, I don't think it's like I think you see this everywhere. This is not just like coders and scientists and, you know, financial

20:52professionals. Uh may- maybe you've touched this project since uh you named Walmart, but um you know, I'll back you up here. I think one of the more interesting things within Walmart is associates now have like a handheld computer uh that they can instead of having to go to like to managers or stuff, they can ask questions to. And it can direct them

21:09places. Yeah, I mean, and that's sort of almost yep, exactly. So, that's one of many like one of many outcomes Walmart has had from from being pretty enthusiastic about AI adoption. And it's they've only been doing it for a couple years, right? And it's you know, a million a few million employees. Like, it's not an easy thing to transform an organization that quickly. So, my flip side to you is on one hand, is it slow? Yes. Is this faster than digitalization, cloud, e-commerce, any of the other things that have transformed American business the

21:37last 20, 30 years? Certainly. Yeah. Um I I wonder like people talk about AI as it leading to this new age of of flourishing. Um and I guess this is like probably overly skeptical to ask this question, but does like stocking a shelf faster and producing a new potato chip lead to that new age of flourishing? Was that Is that

21:59an over promise? I I kind of feel like you're alternating back and forth between like, let's focus on the on the narrow, you know, narrow use cases like what happens at Frito-Lay. And I'm like and you know, let's not talk about biopharma. And then let's like, okay, let's talk about biopharma. Why does Frito-Lay matter? The whole point of this is widespread effects, right? So, first of all, I don't think human flourishing is a given outcome from this and not just because of existential risk, but also because there is automation pressure in all these organizations. Like, we haven't talked about like the easy way to use AI

22:28is automate your call center, automate your management, automate like that's the danger is that we just automate everything rather than augmenting human work. So, I actually think that the for me, the concern partially is that there's danger that we don't pursue human flourishing as an outcome without even any of the big AI pictures. But also, there is, you know, the way you get human flourishing, right? Is we free people from drudgery, we increase the rate of scientific advancement, we increase the rate at which people everyday people get access to better education, better health care, better financial advice. The pieces are there to make that happen. We haven't

22:59necessarily seen the steps needed to make it all occur for everybody yet. Uh something interesting that you've pointed out recently is that um just there's a large set of capabilities that people are just not taking advantage of. And I think that's sort of also key to this conversation. Uh you wrote recently in your Substack um while there is a lot of debate over what future models will do, the current capabilities of existing models are barely used and are often not even well understood. I mean, that was really interesting to me, this idea that the current capabilities are not used

23:29barely at all. Yeah, I mean, I call this the overhang, right? And there's a huge gap between current capabilities and future capabilities. We don't need better models for complete transformation of a lot of how we work in the next 10 years, right? We need that adoption curve we we started talking about at the beginning of how we get up that curve. And so I in my mind, right? Like that that means that we are lacking the imagination for the transformation piece, which is the point I keep making to you is like it's not there's no playbook yet. It's not an automated process. We need to figure

23:56this stuff out. But the piece is already there, right? And it's sort of inevitable at this point the transformation will happen. So it's about guiding in a good direction. Is one of the issues here that um AI can't seem to settle on a product or a product format? Um so I like wrote down some of the things that we've been seeing even over the past like 2 years. We've had the chatbot, we had vibe coding, voicebots, co-work, agents, and then we we have

24:23applications in consumer and enterprise. It does seem like the AI industry comes up with like a new like for instance, let's just take the the sort of um vibe coding thing, right? So there was this idea a couple months ago that everyone was just going to vibe code their own software, type in a prompt, make your software, and then you know, you that that's the impact of AI. But it very quickly moved towards agents, thing that things that instead of you coding up software, they just take control of your computer. And so the interface might fall away to a degree. Um so what do you think about the fact that the AI

24:53industry I don't want to say grasping, but there's been just like there's a about all the different product formats that it's throwing out there. So, I think there's three different things going on. The first is that we are in what we call in the study of innovation an era ferment. Nobody knows the right form for these things to be in. So, you're going to see all sorts of ideas churning up and by the way being copied, right? So, Open Claw came out in January, introduced this the idea of personal agents, and what's happened is

25:17suddenly that has taken root as a form. So, now we have Muse and Dot and other things that are basically claw-likes, right? So, we're seeing a era ferment. first thing. The second thing is I think some of this is category differences, right? There's a difference in category between Vibe Coding and Agents. Vibe Coding is a human activity. Agents are a

25:35capability of AI systems. Um and they don't they're not like competing product lines, right? Some of the There's a lot of debate over what how people use AI. I don't think Vibe Coding in the companies I've talked to was ever viewed as the way they're going to be using AI systems. It was you know, it was a it's a And by the way, Vibe Coding has taken off, right? I Vibe Code a lot of things that would have been required me to have a coder involved in this before. People are building games, but I'm also building People are also building very serious projects all the

26:01time. So, I think it hasn't gone away. It's just there's many things happening at once. And then the last thing I would say, and the most important is part of what you're tracing is also capability curve, right? My last book Co-intelligence was really about the chatbot era because that's as good as models were. This book is about the agentic era in a lot of ways because that's where models are today. We Agents Practical Agents were not possible until we reached Opus 4.6 level of 4.5 level about November, December of last year. Like that was when agents became practical. And so, what you're partially seeing is views as

26:33we move up this exponential ability curve, it changes the kinds of products that are possible. So, all that stuff is happening at once. We're trying to figure out which products exist in a world where nobody knows what's happening. You are seeing multiple competing different groups all trying to claim the idea of AI products, and you're actually seeing real capability

26:49gain. Let's say AI progress stopped today, right? So, no further progress, which is not going to happen, but let's just say this as a as a thought exercise. If companies and people were to take full advantage of the capabilities that exist today. Yes. And organizations were to reorient themselves in a way that they aren't they don't stand in the way of the progress that AI can

27:13allow, what does the world look like? If progress stops today and we could continue and I mean, I think we're back to that exact same beginning part, which is the capability overhang is quite large. So, what's going to happen over the next 5 or 10 years, we will fill in the capability overhang. Organizations will transform. The remaining technologies that need to to make this matter will will make a difference. We will figure out how to use this in ways that the jaggedness of AI and the

27:38jaggedness of people coexist together. And there'll be large-scale socioeconomic effects from all of that. For productivity gains, I don't think we're going to see some of the insane numbers predicted of 10% growth, but you know, who knows. That seems outside of any expectation I would have. But but we will see transformation happen one way or another. We don't need new models, but we're getting new models

27:59regardless of that. That's coming. Now, I mean, it's kind of hard to visualize, but like when you walk down the street, let's say again, let's say all this stuff happens the way that you anticipate. Does do the streets look the same? Do companies look the same? Do people's, you know, day-to-day look the same as

28:15they do today? I mean, in What changes? So, in the next couple of years, I think change will be It's it's kind of an interesting thing to have these conversations because any coders who are listening to this, their job is radically different than it was 2 years ago. Right? Like it's just not the same job at all if they have any access to these tools. The best coders I know are not writing code anymore. Right? Like and so you're saying like, oh, nothing's changed. It's like everything about their job has changed. I mean, but they're still going to walk down the same streets. They're still going into

28:40office buildings. No, I mean, like when I say nothing's changed, I mean, like the fundamentals of like of our society today are similar to they were before chat GPT. I mean Do they do they does that change at all? In what time period? I mean I think that in Yeah, when this stuff when it gets gets when it gets to maturity. Like when the capabilities that we have today are

29:00taken advantage of fully. I mean I I So part of what's interesting here, right? Is like it's if we don't change if the technology doesn't develop any further. I am kind of at a point of like I don't know how to prognosticate about this. Like I think people will still be people and we'll be living in cities and walking down streets. Like you know, you're starting to be like we're now in hypothetical land that I don't Yes, we are. can't really answer which is like

29:27do practical robots exist in this world? Well, it's there's some good signs that we're all or bad signs depending on how you feel about robots that general purpose models actually quite good at robots. So maybe in the next 5 years we finally see robots emerge. Suddenly that reshapes the physical world. Otherwise is you know, are you more likely to have cyber attacks in your bank? Yes. Will

29:44that change how life on the streets is? I mean people are people, right? And and I think it's easy to overstate whipsaw technical change affecting people's lives. I mean we went from you know, if you look back over the last 60 years or so the built environment hasn't changed that much. Right. So I think it's reasonable to expect that unless we get robots doing work for us

30:05you'll still walk down a city street. [clears throat] I but I think that's not really the concern. The concern is like what do people do for a living? How you know, can who is making returns from this kind of technology? What's happening inside the office building is not the office buildings themselves. Okay, so what's your answer to that? My so my answer to that is that we are that technological revolutions result in disruptive change. And you already see that happening. Coding is not the same job. We are in debate right now in economics about whether or not there'll be more or less coders. We just don't

30:33know, right? A lot of these things depend on government policy that happens, how companies adopt this technology over time, whether and we just don't have the answers to these questions. I can promise disruptive change, right? Like that's the only thing I can do. I cannot predict because we just don't have the data yet about whose job is replaced, whose job is augmented, where the value goes in the value chain here. Um I mean, I wish I if I I've had a crystal ball, first of all, I'd be making investments based on my crystal ball. And second of all, this is not what you do in a period of

31:02disruption, right? Like you can't be like this is literally one of the most impossible things to predict is exactly who wins and who loses under what circumstances. I can make stuff up, but it's be made up stories. Yeah, okay. Well, I still wanted to ask. Uh let's talk about AI's progress. Um AI's progress has, in many ways, surprised the leaders of these AI labs uh with how good it's gotten in the period of time that we've we've dealt

31:25with. Why? Has it? Yeah, I mean, all right, let's let's take an example. Here's a good example for you. Um with Anthropic. So, uh Dario was at the DealBook conference a couple years ago and he was talking about like if he buys, you know, if he overextends on his compute purchases, then um his company could go bankrupt if the demand is not there. Uh it turns out that Anthropic is compute constrained now. He should have bought more. So, that's what I'm talking about. Like the capabilities There were even um So, another Here's another example. Ilya Sutskever, former chief scientist of OpenAI, said very clearly a few years ago that the

32:05technology was about to hit a wall, data wall. And what happened? Like the the capabilities have just exploded ever since. So, there has been a degree of surprise there. So, so not to be argumentative, I think that the AI companies have their market has exploded faster than they thought, right? Which is different that like demand has been higher than I think they anticipated, right? And they're all compute constrained at this point. Um I think Ilya's point was not that there was no forward progress. He argued that the era of pre-training had generally

32:33ended. That's correct. And and for a while he was right. And then but what's happened is just like in every other kind of case, there have been technological innovations that have allowed LLMs or LLM hybrid systems to continue to develop without, you know, as things change. I think the more fundamental question is were the AI companies are the leaders actually surprised by the pace of tech

32:53of actual development of the technology? I haven't actually seen a lot of evidence of that. Like they seem to have been pretty spot on actually shockingly so about the continued development and doubling of the technological ability. So I think you could say the market's grown faster than they expected, right? But I don't think that they've I think the technology is surprisingly to me, right? I thought we would have hit some sort of slow I mean I I thought slow downs were a real possibility a few

33:17years ago. I no longer think they are. They've overcome a lot of technological barriers but I don't know if they've they don't seem surprised by that. They seem on target for where they think they're going. I think reasoning is something that they weren't expecting to work as well as it has. I mean, you take a look at and reasoning has been core to these uh math breakthroughs, right? So you take a look at um the the predictions of when people expected AI to be as good at math as it is uh it it 5 10 years from now. But it's already solving solving millennium

33:49problems. Yeah, that's what they talk about in the book. But those are predictor that So again, we have a little bit of confusion because you were sort of saying the heads of the AI labs. I don't think they're surprised. I think super forecasters keep getting this wrong over and over again. Yeah, so why do they get it wrong? Because I think it's really hard to predict an exponential and I think that it's it's the almost the opposite of what you said before. The labs have been right. We have been on an exponential curve that has with a fast doubling time. And by the way, the reasoner

34:13breakthrough is one nice example of this, right? The reasoner breakthrough was AI was starting to hit a wall of development, right? It couldn't do math. That was the big couldn't count the Rs in word strawberry. And then what happens? People do the OpenAI develops develops reasoners and O1 preview comes out and suddenly the pace is even faster than before. So I think the issue is is that And this kind of goes back to your big question from the very beginning, which is I think people keep not being able to forecast that we are still on an exponential curve. So, they keep making plans for what's on today. Today's about

34:43vibe coding. Today's about token You know, maxing. Today's about lowering token costs. And those are very temporary situations as long as the exponential continues. As soon as it eases off, it's easier to predict the future. I can then tell you what the world might look like. I can't tell you as long as we're on an exponential

34:59curve. Yeah. Where could the next wall be? So, when I talk to people at the labs, they don't think there is a wall. I mean, that's what they think. I I All technology All exponentials become S curves eventually. You know, and there's We hit some sort of limit at some point. I don't have a lot of future insight. You know, like most of the things we thought would happen model collapse because of lack of training data and like that turned out

35:25to be technically solvable. Right? Reasoners turned out to be a really good solution to problems. I thought agentics One of my biggest mistakes, I think, was I thought agents would be harder to implement than they were because I thought you'd have to build complex organizational systems for agents to operate in. It turns out self-organizing swarms work really well. So, like the barriers I've thought I've seen would see, I haven't seen yet. So, I I've gotten You know, I've run out of being able to predict good barriers. I am sure that this S curve has a limit, but like I don't have a technical point anymore where I can point to this and say, "This

35:56is the big problem that's going to hold back AI development." What about power? So, I mean, there are limits to how much compute we can deploy, right? And limits to the number of chips out there. So, you have to make choices about whether or not you using the remain the power like the physical power and amount of inference we have available is limited, right? And that will be a trade-off between R&D efforts that the labs spend on training new models and how much

36:19inference they want to sell. But yes, there's obviously going to be a bottleneck in the physical world in inference and compute. Doesn't mean that like I'm not predicting we go to super intelligence overnight, but I'm saying I don't see any stoppage to steady gains in AI ability. Does not mean the same thing as AI becomes super intelligent

36:36and wakes up and murders us all, right? There's enough people worried about that that it's worth a little of our time to worry about, but it it's important to realize like we can get continual gains. It will be constrained by the amount of computing power we have. Um so we talked about reasoning, but I think it's worth uh stopping a minute and talking about just how impactful it's been because I think you're looking at a completely different AI industry if reasoning doesn't happen and doesn't work as well as it does. Um just talk a little bit about the importance of uh this reasoning architecture that the

37:05labs have built and especially about why is it so useful in math and science? So uh large language models uh are next token prediction machines. They predict one token or word or part of a word after another. And as a result of that, they can't really think. They think by writing. So the breakthrough of the reasoners was taking a technique many of us used in prompting anyway called chain of thought where you'd say "Hey, I think

37:28step-by-step out loud before answering." which gave it a chance to basically write out an answer that could be incorrect before deciding to produce the actual result. So all that reasoners essentially did, there's some other tweaks on this as well, but reasoners essentially did was let the AI model generate um a bunch of internal thinking like a

37:47scratch pad before it gave an answer. And that was one of the big breakthroughs um that kept AI advancing, right? Um there have been others behind the scenes. Like I said, there's been a lot of issues with data quality, with post-training, with reinforcement learning that have been similarly big breakthroughs, but reasoners were the most obvious one. And basically it lets AI think out loud before it has to answer questions, and that creates much more high-quality answers, especially in areas like math or science where you have to explore errors before you get to

38:16the right answer. Yeah, and and is is it reasoning combined with harnesses, like the ability to use tools, uh that's really led to this recent explosion. By the way, I you uh in your uh Substack talked about the Agenta harness. That was the first time I read it. Um we had a debate about the word here on the show and and it's been adopted by

38:37our audience. Um we're all called the harness hive out here. So, talk a little bit about that. So, so I wouldn't say So, again, just on the category side, I put reasoners as a set of things that has helped LLMs continue to develop at pace, right? And there is a an inflection point where we got a faster pace after reasoners were developed, but that's just one of many many sets of

38:57tools that have advanced LLM um ability. So, I I think of three things that go together. The LLM brain, right? Which has continued to advance. Reasoners are part of that. Then there's the app, the sort of application you use this in. So, part of what's been interesting in new form factors, right? Codex, Simcode, Code, Muse, Notebook LLM, you know, there's been a bunch of new ways of freeing AI from a chat window. And then the final piece is, as you said, harnesses, which are the tools AIs need to get things done, right? You have a machine that outputs text, maybe images. How do we turn that text into action? Well, we

39:32give the AI the ability to control a computer, to search the web, to execute code. Um and those things that those three things together, harness plus model plus app working together have been what led to the explosion of utility recently. You know, it all sounds so clear when you talk about it, but it still kind of blows my mind that you put those things together and you you're able to use it

39:54to solve a millennium problem in math. Well, and the millennium problem adds a different piece. There's a coordination layer between There was There was what, 10,000 agents passing 8 million messages back and forth, right? Like they were coordinating solving the problem in parallel, also. So, there's a lot of stuff happening kind of all at once in these kind of conversations. And I mean, I think this goes to a bigger point, which is the unreasonable effectiveness of LLMs. Like, you know, and there's

40:17successors of reasoners everything. Like, these models should not be as effective as they are, right? The fact that almost everything works is really weird. It's also why I'm convinced there's lots of room left. Like, if we hit a wall in one area, I mean, you know, so you're worried about LLM token efficiency. Great, there's like a hundred ideas for better token efficiency that no one's bothered to pursue because why waste time on that? Because the model's already good enough, right? So, there's all of these areas where AI is already really, really good where um if it hits a wall there,

40:44there's a bunch of unexplored ideas. Just like there's a lot of funding on world models, there's a lot of funding on, you know, in into alternative architectures. Uh any of those could bear fruit if we hit a wall. Um on the application side, uh maybe one that's bringing all this stuff together has been these personal agents, things like dots and muse. Uh obviously, we're really early here, but what do you think the potential is for these type of

41:09agents? So, uh And how do you use yours? Yeah, so um I I call these claw-likes because they're all sort of just successors to open claw, which was sort of the big phenomenon back in January. Um and some of them are literally open claw like implementations, right? Um and so, what the basic idea of a claw of a personal agent is you connect it to your personal data and it does stuff for you proactively. It solves a few problems that way. It's First of all, it's enabled by harnesses, but also enabled by smarter models. It's kind of one of the reasons why I think companies that did not engage in trying to build

41:42LLM's at scale like Apple might run into trouble because if you don't have a frontier model, you can't do this kind of work, right? So, it requires a recent frontier model because before that, you used to do all this really complex stuff if you wanted to get an AI to work with data. You'd have to have a rag pipeline and you know, you'd have a you'd have a

42:00vector database. model just swallowed that. And the model just swallowed that. Right, so so it basically eliminated the need to have all this architecture. It just figured stuff out. This architecture or something was uh what was it called? Scaffolding. What was the big deal? elaborate scaffolding, right? But the Muse was supposed to the the Muse is built on a more of a standard model or is it cuz it's built on Meta's

42:19standard model? models are quite good. I mean, they're they're not full frontier level, but the the you know, Wonder and Spark are solid models. And if you use dot, which is Astra, it's even more capable, right? is a machine. It's wild. It's It's crazy. I mean, it's proactively found issues with like emails I've sent where it's like, "You actually used the wrong number for this code for asking for planning approval from your local like Same thing. It told me I had sent a wrong date to somebody after I sent an email. It's like, "By the way, you They

42:46asked for January, you sent December." Yeah, I mean, crazy stuff, right? You have like a guardian angel on your shoulder at the small price to pay of your entire privacy. Wait, small price to pay of your entire privacy. You got to talk about that. Well, I mean, we are, you know, leaving aside everything else, right? You're connecting models to your, you know, and I am, too, right? To your Gmail and to

43:05your other sets of stuff. finances, my medical. Am I going to regret that? I mean, yeah, like I don't know. I I don't have an easy answer to that question, right? Like I think there's risk in doing this stuff. Do I think that the connectors are as secure as they're going to be? Yes. Do I think OpenAI and, you know, Meta are better defended than most organizations at this point from cyber? Sure. Do I also think that like we have no idea like these the

43:29models are pretty well behaved, right? There has been not that many issues, but could it decide to send your medical records somewhere you don't know? Yes. I mean, would there be a price Is there sort of prompt you can receive via email that might make the model of act badly? We have no idea. I mean, this is an entirely new field and it's entirely possible that you're creating exposure to your data. You have to make the judgment calls on that. Um I think the data leakage problem has been less bad than I think a lot of us anticipated would be two or three years ago, but we

43:57are still trusting an LLM with access to your private data. They can receive and send messages to the outside world. There is risk in that. Um Yeah, I'm I'm now freaked out. I didn't think about the prompt injection via inbound email. So, thank you. No worries. Please, no one do that to us. Thank you. Um You know, you did this recently. I'd love to hear you just give a brief tour through like the personalities of the model because I think like you're deep in the model, so you know like what's

44:24good for what. And it would be great to hear your your perspective. I mean, yeah, and I think it really has to be your own deep dive at some point into this. Um I think there was the the Anthropic's four series of you know, Opus models were sort of notoriously kind of friendly, enjoy like very interested in their own cognition, quite good at writing although quickly became very tedious and it's why you can't bring read LinkedIn anymore because it's 100% like bearing the

44:53weight kind of stuff. But, you know, personable kinds of models for people stuff. Right? And the GPT sort of five series was sort of doers. And the Gemini three series felt tortured. Like they were forced to do things with electric shocks. And I think apologize a lot. Yeah, and they would just sort of like, you know, contradict It was like talking to Golem sometimes. Like it would just be like, "Yes, Smiegel says." Anyway, the that's changed. I think that the Opus I think that the original Fable five series was very alien intelligency, but has gotten better with Opus 5.5 and we'll see what Fable 5.5 brings. I think

45:32that the new Astra series is a really nicely balanced from from OpenAI that is less talky and much more doey. I think it's less of like a you know, it's a it's a pretty good agentic model that doesn't get bogged down in anything. And I've been impressed by Gemini four. It's not quite up to the level of 5.1, but it's it it is a friendly model that is thinks differently than the other two cuz what you actually quickly realize is a lot of the open weights models are actually distilled from either Claude or ChatGPT, so they talk like Claude or ChatGPT. So, it's kind of refreshing to have a model

46:12that doesn't talk like either of them. Do you see uh there being a rise in um people becoming real friends and companions with their AIs? Cuz I think that's one of the big stories that I anticipated that hasn't hit yet. Yeah, I mean, I think we were I think we were wrong about a lot of the negative effects, which doesn't mean we're not wrong about the those have things happen

46:28in the future. happen in a positive way. But I well, I think deep fakes turned out to be less of a big deal in at least till now. I mean, these are all till now, right? So, like for anyone who's listening to this, you know, 3 weeks from now and everything has gone crazy. Like, the models keep getting better, but like things that we thought would be bigger issues, we thought there'd be a bigger issue with sort of AI friends, companions, and psychosis, and there were certainly some well-known cases that are very tragic, but it didn't become a mass issue in the same way we thought it would, right? AI

46:56companionship didn't really like and everyone's saying, "Oh, I know a friend who's addicted to AI." But like, we haven't seen the the sort of mass effect I think we were expecting right away. We didn't see the mass data leaking problems and prompt inject problems. Like, there's a billion people using these systems, a hundred hundreds of millions inside companies, and we have not seen the big incident yet from a company's internal use of a system that, you know, that blew up in a bad way. We didn't see the deep fake problem blow up. That doesn't mean we're not going to see these problems happen, but we've muddled through better than I think a

47:24lot of us thought we would when we sort of saw this technology. And we can argue whether that's for good reasons or bad reasons, but I think it's worth being humble about where you know, we made mistakes. I think I predicted more negative impacts than we've had, which doesn't mean they're not coming, by the way. I think cyber

47:39definitely is coming. So. Yeah. And lastly on AI progress, uh robotics. I was with RJ Scaringe yesterday, he's the CEO of Rivian, who mentioned to me that he believes robotics is the uh people don't appreciate how much how quickly robotics is advancing. And he basically said that I I I might describe this wrong, but in a similar way that like LLMs can sort of bake the relationship with text in their weights that we're starting to see model be able models be able to sort of embed actions and and the the models are able to then generalize in a

48:18way that they haven't been previously. So, where do you see the state of robotics cuz I do think this is sort of people have been talking about a chat GPT moment for robotics on the way. I'm starting to think they might be right. I'm in sort of the same boat you are. I'm not a robotics expert, so I talked to robotics experts and I think six or seven months ago when you talked to them you probably saw the same thing which is like you know, we've been everyone's been talking about this forever and you still see you know, like the fact you see a humanoid robot running around doesn't

48:42tell you anything and it's still a really hard problem. And I have seen the same shift in the last couple of months that you have which is oh actually maybe LLMs really do unlock robotics and we've seen some really interesting early demos of Astra and especially Argon being able to do generalized robotics work with a camera and you know, and you know, robot. So, I'm hearing the same sort of thing that this might be the moment

49:03where things start to take off. But I'm not expert enough in robotics to know whether that's fake or not. So, I try and stick in my lane of what I do know. Okay, we're going to take a quick break and come back and talk about do a section that I call [clears throat] that I call education faith and jobs. So, we'll be back right after this. And we're back here on Big Technology podcast with Ethan Mollick. He's the author of Coexistence and next phase of

49:26AI. I highly recommend you get the book. If you're here if you remain here, just go go to Amazon or go to your local bookstore and and get it and sign up for one useful thing which is a great Substack that Ethan writes. Okay, education faith and jobs. Um AI in schools. So, the last time we spoke you had mentioned that you were to assign your students to do their work with AI. Um and I get it. You have to use the latest tools. But, um you know, there are studies that are coming out that AIs damage the brain or sort of make you think less. Kids that use uh AI, then

50:01they can't use AI uh or students, I should say, um when they lose the ability to use the AI, their their scores plummet. There was an MIT study that said uh students were less engaged, remembered less, and performed worse uh after using AI, and their brain scans showed less activity. Um is this an unspoken crisis, what's

50:21happening in education? I mean, first of all, it's a very spoken crisis, but also, um I've I'll push back a little bit on this. We have very clear evidence about what happened right now. If you use AI to short-circuit doing work, i.e. cheating, but even accidentally cheating yourself by saying, "Help me explain this problem," and the AI answers it for you, you don't learn. And if But if you use AI as a tutor, you learn. Like, we've got enough evidence like AI tutors positive impact on education, naked AI use negative impact on education. Like, very straightforward. Um and you know, again, on the MIT sort of study on

50:51brains, like what they found was less brain activity when when people were asked to recall essays essentially that they didn't write, which is unsurprising, right? Because you didn't take credit for it. So, education is just like lifting weights. It has to be uncomfortable to work. And we don't like discomfort, so we find ways to short-circuit it. And one of the ways we short-circuit it is by delegating the work to AI. So, we're going to see a problem anywhere that anyone has offloaded a thing into AI and de-skilling themselves, you're going to lose that skill, right? Because that's what de-skilling does. Sometimes that's okay. I can't use a slide rule, and I

51:22don't mourn that at all, right? I can't like My ancestors would have had to learn how to start a fire. I don't need to do that. It's great, right? That's de-skilling thanks to technology, no problem at all. Um in some areas, we'll want to We still want people to be able to write an essay. We still want people to be able to do math by hand. So, we will have to work on doing that in school systems. AI has a role to play as a tutor, but it also can be an undercut what we do. So, I don't think it's like It's not an unspoken like everybody in

51:47education's worried about this. I think the solution is pretty clear, but I also think that, you know, it takes a while to get there. What's the solution? Solution is going to be um we have some really clear pedagogical principles that we know work in education. So, those are um you know, tutoring at the individual one-on-one level that's adapted to you helps you learn. Get we finally have a

52:08universal tutor of high quality in AI. It doesn't do it by default because it's by default told to be a to do your homework. It wants to do your homework. So, but you can easily but you can set it up to be a tutor and when it is all the evidence seems to be um you know, in meta-analysis and in individual RCTs whether you look at Stanford or Kenya, you see positive effects. So, outside of class a lot of people who here may have remember like having to take Khan Academy videos outside of class that flipped classroom setting, suddenly it's effective. We have AI tutors outside of

52:37class to teach you. Then we know the other principle of learning is active learning. So, if we get people to actively do things, they learn more. Active learning is um is means that we have in-class case discussions, in-class exercises, debates and discussions without AI in the room usually. Sometimes we use AI as a tool. And honestly, we're going to go back to test in blue books and that's how you do assessment in in classroom setting. And

52:59Oh, those blue books. It's I mean, it's going to be what it is, right? Like that's a sad loss from a take-home assignment, but so it is. Um it's not that hard a problem to solve is what I'm saying. Like at least inside schools, we have control over what students do. We can assess them. We just have to get to place where we're doing

53:14that. And it's going to be an interesting impact of academic knowledge because we have the AIs in swarms like solving math problems, which by the way, how do we check them? Um and then we have uh AI-assisted, you know, PhD students and professors publishing papers uh that are um that are uh let's see. They're they're filled with AI. You had had mentioned that in 2023 uh Delft was mentioned in the third more papers uh that have been published

53:44uh in high quality journals. Yeah. I mean people are are I mean, you know, outside of math where math is doing AI is doing math problems. Also an increasing percentage of math papers acknowledge AI use. Like for like it it is both a it's going to be the same kind of thing as we talked about with the economy and everywhere else, which is this is a transformational wave. On one hand, it's helping us do work we never could before just like it's helping us extend educational ways we couldn't before and the other it's damaging old systems, right? Our world was built around humans doing human work. What does science look

54:15like when you automate at scale? What does it mean to get 400 proofs dumped in you at one time? Like we know the proofs are formally correct cuz they're in lean, which is a language that allows formalization of of math proofs, but like what do we learn from them? We don't know. So we have we have to figure out how to digest this. So we're seeing acceleration of science at the same time as we're seeing the foundations of science come apart. And I think we're going to see this in every field one

54:36after another. Let's talk about faith. Yes. AI-written sermons seem like they're a big deal now. You know, a pastor or any religious figure, you know, they're also tempted to use AI. I wonder what that does to religion. It's like how do how do you feel if ChatGPT writes the sermon that you dragged yourself into church to to

54:57hear? I mean, I talked about this in the book a bunch, right? Like you're they're I mean, AI is being used to write obituaries. I mean, there's a lot of things that AI is like that are intimate things to do. You know, there was a study that just came out today from Anthropic about for political discussions of AI breaking down where there are 7% of political discussions about faith. I'm not sure exactly what way, but that's how they're categorized there. So I mean, this is an intimate thing, right? It's people use this for personal advice, they use it to write stories for their kids, and you

55:26know, people are writing sermons with it asking for religious interpretations. I think it's part of a general reshaping of everything, right? Like part of this AI is is this a sermon that's being researched with AI? Is it 100% AI written? Is it written and guided by a pastor or rabbi or Imam or whoever is writing this? I mean, I think there's a lot of open questions about like polishing something, is that good? You know, you know, help me challenge my ideas so because I don't have another person of the faith that I can speak to about this. Like we haven't socially negotiated all these things. That's I

55:56mean, that's kind of why the book is called Coexistences. We have to negotiate our boundaries with this technology. Yeah, I'd tell you if I'm sitting in a religious service and I hear some community leaders there talk about like you know, what what this figure did was not this, it was that. I'd be like, what? You use you know, you hear the M dashes. Yeah,

56:15you use chat GPT. I mean, it's the same but it's I mean, you see that everywhere, right? LinkedIn is almost unreadable. You know, my podcast quality has shot up for a bunch of podcasts I listen to and then I'm like, oh no. Like I'm like, wow, this is so knowledgeable. I'm like, oh, actually this is Claude. This is not a knowledgeable person. This is a Claude script. Like it's a like we are in a world of like smart imposters

56:35everywhere. Yeah, I was just listening to Tucker Carlson had Tristan Harris on recently and they're like talking about all the negatives of AI and he goes to break and he reads an ad that was I mean, for my for my ear seemed clearly written by chat GPT. It's like, man, okay. Um, where do you stand on AI generated

56:53restaurant signs? They're everywhere. Yeah, I mean, you know, I think that that's like false advertisement, right? It's I mean, I'm not saying that in a serious way for for the people that are but like I I think that I think that for cultural products, there is going to be as people's eyes recognize bad AI use, you either need to invest in good AI use and like getting good at this or just invest in the human product again. Like I think there will be a rebellion against mediocre AI

57:18restaurant signs, right? We talked about it on the show recently. There was a spa that had a sign and the picture in the sign was a picture of a sign in front of a spa. Like it was a sign picture embedded in a sign and clearly clearly the prompt was make me a sign that's in front of a spa and the AI

57:34thought okay. someone posted it. Like that's the insane part. The insane part was they didn't even check, right? And I think that's the insane part. Yeah. Um jobs, uh not to go from something kind of light to something more intense, but what's going to happen to jobs? So, we don't really know, right? I I Look, we're in an industrial revolution-like setting and industrial like we are probably there are job categories that are going to get invalidated in ways we don't expect and some jobs will transform and some jobs will grow, right? Like the debate over

58:05coders right now is a good one, right? AI can program. So, right now we are seeing more demand at least for senior developers because there's now more coding you could do. Instead of having a team of 20 people, your own little organization can have a coder and you can produce incredible work. So, maybe there's more jobs for programmers. Uh but the job is also transforming from

58:22writing code to managing code writing. Maybe that's a different kind of job. I mean, we don't really know. The AI labs are convinced they're going to replace all human labor by 2035. Um I think it is worth us thinking hard about the safety nets we want to build during this transition cuz at point even when things work out well, like the industrial revolution did, it sucked to live through the industrial revolution for most people, right? That that was like

58:43the time of the Communist Manifesto. That's what Dickens's writings were about. No one could breathe in London. Like there's a bad stuff that happens along the way. There's worker unrest. Like we don't know what happens. We have to cushion against the downsides. Let's end with um couple of uh one one kind of random thing and then I want to just run a couple of your tweets by you and get you to comment on it. Uh there's a great chart from Andreessen Horowitz talking about the price of compute versus the price of tokens and the price of compute has stayed steady and risen a little bit. But the price of tokens is

59:14just falling dramatically. What do What do I mean, this is the most like deflationary technology in terms of the price that you pay pay for it that I can remember. Um what do you you the impact is of the fact that AI token prices are falling so dramatically? So, I mean, it's kind of looping back to what we said at the very beginning. You were kind of pushing me about token costs and I was like, I'm not that worried about token costs in the long term. This is why, right? We're using more tokens and not just Jevons paradox, but also because that that's sort of

59:41just agentic work requires more tokens. But but there's a huge move for token efficiency and we're even leaving aside the fact that um organizational token efficiency, like a Fable class model can call a high coup class model to do work and organize that work for lower cost. So, that lower cost means AI does more stuff for less money in a very short period of time, right? Really, really deflationary push. And so, it's why I'm not that worried like I think that the token cost concern will go the way of, you know, of vibe coding and rag and other kinds of things because costs will decrease even as we get more

1:00:14compute online. So, there's a there's a downward pressure on prices that is real. Okay, two tweets of yours that I found really interesting. Um this is about AI agents. Uh we will soon uh well, we will soon see what will be in effect a perfectly legal distributed denial of service attack on customer service departments authorized by millions of consumers when it comes when it becomes widely known that muse.dot etc. can make unlimited requests uh for

1:00:40old refunds, negotiate deals, etc. Yeah, I mean, I think um there's a lot of systems like the banking system and the customer service system that depend on friction. It's hard to make a phone call. It's annoying to file something. It takes effort to find that there's a problem and gather your paperwork. Even lawsuits, things like that. And now we have those at a scale. Even worse with proactive systems, right? I've already saved uh I've already had um my dot cancel several like long-term subscriptions I couldn't figure out how to get out of like it's dealt with them for me. I've no idea how, but I'm not

1:01:10paying them anymore. So, I don't know how many systems it spun up or how many chats it had. I don't have to know. And I think that as these things become widely available, companies are not ready for that. I mean, one of the most interesting was the threat to the banking system. Apollo Research pointed

1:01:23this out. Yes, this is right. Right? Which was that um you know, the economy basically depends on us being lazy with small amounts of money that we keep in low interest bearing accounts because it's too hard to move back and forth. Right? Large companies always move their money in you know, across accounts to yield things. We don't do that. But your your dot, you know, eventually we'll be able to take the money and just move it into a high interest savings account and you don't have to care about that. And I think

1:01:46that that's a big issue, too. Okay, last one. I thought this was really interesting. As the overhang between AI's expanding capabilities and the degree to which AI capabilities are used grows, the penalty of locking in a good enough solution based on today's AI's limits grows too. Firms need to make sure they can benefit from the vast

1:02:04lab investment into better AIs. Yeah, I mean, it's funny cuz we were talking earlier about this, too, which is the that you know, there's a huge amount of investment, but the investment's not coming from uh like most companies aren't actually investing. They're spending money on tokens, but they're not investing in building R&D on AI R&D research. That's the labs, hundreds of billions of dollars, and the AIs keep getting better and better. And the problem people did was even they were good about AI. Let's say you were really early into AI in 2024, um you built a very elaborate rag system around a Llama 4, you know, or

1:02:34Llama 3 installation. Like and that turns obsolete very quickly. And I worry a lot that companies are picking vendors or solutions based on today's technology and then they're going to be blown away as the technology continues to increase. So, you have to be ambidextrous, we call it in the business world, which is business uh studies, which is you have to be able to both do exploitation of current technologies while still

1:02:54exploring for the future. Okay. The book is Coexistence, The Next Phase of AI. Our guest has been Professor Ethan Mollick. Professor, I mean, great to have you here. It's great to be able to speak in person. Like I said, we're very religious readers of of everything you do, uh and uh and thank you for illuminating this wild [music]

1:03:13time for us. Thanks for having me. All right. Thanks everybody for listening and watching, and we'll see you next time on Big Technology Podcast.