Economist Tyler Cowen (author and professor) and Google DeepMind economist Alex Imas join a16z General Partner Erik Torenberg to debate why rapid AI capability gains have not yet produced extraordinary economic growth. They argue that institutional inertia, physical bottlenecks, and the difficulty of AI agents coordinating like human organizations will likely keep GDP growth in a normal range for years, even as AI transforms specific tasks.
> "Often the faster that some things move, that just means you hit the real bottlenecks more quickly." — Tyler Cowen
> "What you actually need for real economic transformation is not just competence, it's the ability of these agents to coordinate with one another." — Alex Imas
> "If you ask me the question, when will agents be 10% of GDP? I don't think it's soon." — Tyler Cowen
AI capabilities are improving fast. But will the economy move just as quickly? a16z General Partner Erik Torenberg sits down with Google DeepMind economist Alex Imas and economist Tyler Cowen to examine how advances in AI translate into growth, jobs, wages, and productivity.
They debate what it would take to reach double-digit economic growth, why bottlenecks and falling AI costs could slow the impact, and whether frontier lab valuations require much larger productivity gains.
They also explore the “economics of intelligence,” including why the bigger shift may come not from smarter individual agents, but from AI agents learning to coordinate at scale.
Resources:
Follow Alex Imas on X: https://x.com/alexolegimas
Follow Tyler Cowen on X: https://x.com/tylercowen
Follow Erik Torenberg on X: https://x.com/eriktorenberg
Stay Updated:
Find a16z on YouTube: YouTube
Find a16z on X
Find a16z on LinkedIn
Listen to the a16z Show on Spotify
Listen to the a16z Show on Apple Podcasts
Follow our host: https://twitter.com/eriktorenberg
Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
Erik Torenberg
Institutions are not quite running as smoothly as either we might like or we might fear. We have these weird things happening with NVIDIA. If you look at prediction markets for what the Anthropic IPO will go for, it's not flying above $2 trillion. Do you expect the growth to continue or do you think of leveling up?
Alex Imas
Benefits will exceed costs, but it will take some time, and the path will be quite a bit more normal than many people would have you believe. What you actually need for real economic transformation is not just competence, it's the ability of these agents to coordinate with one another. There's this idea of agentic swarms. But if you unpack it, these agents are still very bad at coordinating. They're mostly just doing parallel processing.
Tyler Cowen
Often the faster that some things move, that just means you hit the real bottlenecks more quickly.
Erik Torenberg
If you ask me the question, when will agents be 10% of GDP? I don't think it's soon. What do the two different worlds look like? One in which GDP growth is a bit more normal and another in which it actually becomes double digit. AI models are getting better at an extraordinary rate. So why doesn't the economy look extraordinary yet? In this episode, I sit down with Google DeepMind's Alex Imas and economist Tyler Cowen to ask how rapidly advancing AI translates into economic growth. We unpack why better models don't automatically mean double digit GDP growth. From institutional inertia and regulation, to physical infrastructure and finding productive uses for AI. Alex also argues that the bigger breakthrough may come when AI agents learn to coordinate well enough to outperform human organizations. And that raises the larger question. What would actually have to happen for AI progress to produce an economic takeoff? Alex, Tyler, welcome to the a16z Podcast.
Alex Imas
Happy to be here.
Tyler Cowen
Thanks for having us.
Erik Torenberg
There's a lot to talk about. First, let's talk about some timely topics. The Meter report came out. There's been a lot of conversation. Paul Christiano has talked and tweets. There's a lot of unrest and anxiety around the state of AI right now and RSI. And I'm curious, Alex, what is your reaction to what we are seeing? And where do you stand on some of the proclamations being made right now?
Alex Imas
So, I read it closely. I think it's undeniable that we're having incredible progress and the progress is accelerating. I think anybody who's looking at the model releases, they're becoming more frequent, they're becoming more powerful, the capabilities are going from last November to this November. It's just an unrecognizable world, especially with the difference between agentic AI and the chatbot era. You're seeing what's coming out of companies when you have data coming out on the difference between the chatbot era versus the agentic era. You're seeing real productivity now. With the chatbot, you're using it almost like a fancy version of a Wikipedia.
You're asking questions, it gives you answers. With an agentic system, you're seeing actual work. And that level of work, the ability of these systems to actually do the work, is just progressing so quickly. And you're seeing a lot of hockey sticks, basically. And that's what the Meter report showed is that, look, here's another hockey stick. To tell you the truth, I didn't update too much from the Meter report, because a lot of the data in the Meter report has been around in other spaces. So the S1 statement, we knew about some of the unreleased models. I didn't update too much, but that's because I was kind of already there.
Erik Torenberg
Is your view that the hockey sticks that you're identifying, that they will just sort of continuously go up? At what point might they peter out? How bullish are you from here on some of these hockey sticks maintaining?
Alex Imas
The thing you said, some of these hockey sticks, is the main term. Where are the hockey sticks going to be? How are they connected? In the Meter report, the key phrase was "on well verified, on well specified tasks." The big question is, what about less specified tasks? A lot of work is not well specified. I'm not one of those people who thinks we're safe on the non-well specified side. I think the models are becoming increasingly better on not very well specified tasks as well. But it might be harder to do hockey sticks on those sorts of tasks. There's also a lot of questions of tacit knowledge, Hayekian concerns about the actual impact on the economy.
Tyler and I are both economists, so all of this information I have to sift through about what these hockey sticks mean for economics and what it means for economically valuable tasks. The big questions for me are, with every model iteration, what are the economically valuable tasks that this model is unlocking that the previous model wasn't unlocking? This is captured very imperfectly by the hockey stick. You can get a hockey stick that looks crazy on a benchmark and then not really see a lot of acceleration in the real world. We're not there yet. I think we're seeing a lot of acceleration in the real world, but these are just some of the questions that I have.
Erik Torenberg
Tyler, let's go to you. I feel like you've been a more sober voice in the conversation. Alex has also been sober, but I just mean, have you updated at all based on what's been happening here? What is your view?
Tyler Cowen
Well, often the faster that some things move, that just means you hit the real bottlenecks more quickly. So I don't update that frequently. But this is AI land, and in AI land, the exact dates really matter. So we're recording on November 11th, 2026, and that matters. I would say since August, I've changed my views a fair amount, most of all because of the math problems. Progress in areas where there are not many human bottlenecks, I see much more rapid progress, and the ability to create a new model that is clearly better within something like three weeks, I hadn't thought that would happen as quickly as it had.
So I would date that to a few months ago. But if you mean over the last week, the Meter report, the last two weeks, I'm actually slightly, very slightly more bearish. We have these weird things happening with NVIDIA. I have no prediction, but it will slow down the sector a bit. If you look at prediction markets for what the Anthropic IPO will go for, they bounce up and down, but it's not flying above two trillion. You look at Anthropic's own projected revenue numbers. I forget the exact numbers. I wouldn't say I was underwhelmed, but they're not numbers that you would expect under the super fast takeoff scenarios, and that's for their IPO.
So I'd say I'm like 1% more bearish over the course of the last week than I was a week ago for those reasons. The institutions are not quite running as smoothly as either we might like or we might fear, depending on your point of view.
Erik Torenberg
Are the growth rates of OpenAI and Anthropic, is your base case that over the last two or three years, those growth rates continue? Put aside those specific companies, but I more mean just the leading AI companies, the economic impact that we've seen, whether those two companies maintain the lead. Of course, they're competing with the hyperscalers too. Do you expect the growth to continue or do you think of leveling up?
Tyler Cowen
Well, I was in the Bay Area a few weeks ago and many people were pushing very high revenue numbers on me. And I asked them all the following question. I said, what are the other sectors that are going to grow sufficiently rapidly to justify the extreme revenue numbers you're giving me for Anthropic and OpenAI? And no one had an answer for that question. They didn't even try to give me answers. So I think that means the extreme revenue numbers are false. That doesn't make me bearish on the sector as a whole. I'm still bullish. I think it will be fine. Those are great companies. But at the end of the day, you look at market data, whether it's real interest rates, stock prices, anything.
Market data still look pretty normal. And the whole world investing community is talking about AI. It's not some kind of weird mystery like it was back in 2014 where you could say, I read a LessWrong blog post and now I'm really wealthy because I did. So I trust the judgment of the market more than the judgment of the Bay Area people who don't think that carefully about social interactions. So that's where I stand. It'll change the whole world. Benefits will exceed costs, but it will take some time and the path will be quite a bit more normal than many people would have you believe.
Alex Imas
And just to add to that, I think Tyler's exactly spot on in the fact that you need to start seeing the productivity numbers in the rest of the economy to justify the revenue increases. So at some point, the rest of the economy needs to be willing to pay for that product. And that increasing willingness to pay has to be from increasing productivity in the rest of the economy. And we haven't really seen too much evidence of that. Not to say that it won't happen, but it does need to happen at some point. The other part of that is competition. I think what's starting to happen, particularly in the last few months, is extreme price competition from the frontier labs.
Before, they were just trying to compete on who has the better model. And now that the models are being used by an increasing number of enterprises and they're becoming quite expensive because agentic AI uses way more tokens, companies like OpenAI, Anthropic, Muse is in the game, DeepMind is in the game, are trying to say, no, we're offering a cheaper model that's as good or maybe even a little bit better or maybe a little bit worse. And so you're seeing token prices come down a lot. Ramp has a really nice data set to track these things, and you saw exponential increases in token usage and flat revenue numbers.
Where is that coming from? It's because of price competition. So at the frontier, you're seeing way more price competition than you used to six months ago.
Tyler Cowen
And the very recent elections, I think those are being overplayed. I think a lot of data centers still will proceed, but I think it will slow things down a bit. There's greater likelihood of regulation or somewhat tougher regulation or regulation a bit sooner, or Trump giving on the issue of guardrails. To me, it's not a major force, but again, it's going to make me a modest bit more bearish on the sector in terms of speed of takeoff, revenue numbers, anything else.
Erik Torenberg
By the way, while we're on the topic, did the midterms cause you to update anywhere else, Tyler? Or is thermostatic backlash the thing and we should just expect these pendulum swings often?
Tyler Cowen
Well, most of it I expected and indeed predicted, but Texas truly surprised me. I didn't think Texas would flip the way it did. And maybe I just haven't been to the state enough lately, but that was a shock.
Erik Torenberg
So AI and the economy. You both talked about how you push back strongly against the idea of double digit GDP growth at some point. You're far more sober that it will be tremendously impactful for consumers. It will change a lot of things. But GDP growth, as you said a few minutes ago, might be more normal than we might expect. But before getting into why that is, can you explain what would need to be true for, or what do the two different worlds look like? One in which GDP growth is a bit more normal, and another in which it actually becomes double digit. And Alex, you were alluding to it a little bit. Paint that world to me of what it looks like if it's actually having massive impact on the economy.
Alex Imas
There's a beautiful paper out there that outlines exactly the conditions you need for this singularity type growth by Ben Jones, Chad Jones, I believe Aguillon is on that paper as well. Essentially what you need is a bunch of pretty knife edge conditions to be met, particularly the main one is essentially no real bottlenecks. And the main bottleneck that can't exist is the idea production function. So you can't get decreasing returns to ideas. Basically, you go from the old world where you have this technological breakthrough and then you have decreasing returns to that technological breakthrough.
From that technological breakthrough, you start creating more and more technological breakthroughs. So you need less resources while getting more and more technology and more and more production out of that technology. And so you basically have increasing returns rather than decreasing returns. You can actually estimate this parameter in the model, and if it's greater than one, you're in this double digit growth world. If it's less than one, you're in the standard diminishing returns, single digit growth world. If you look at the data so far, including the Meter report, the data that's out there, and you take that data and you estimate this parameter, it is decidedly less than one.
This could change, and some people say it will change, and when you get the full RSI loop, this is exactly what it's going to change. But right now, you need increasingly more resources in order to eke out increasingly more intelligence out of these models.
Tyler Cowen
I would add to that, you can look at cases where growth has been very high. Not long ago, Guyana grew 40% in a year. That's crazy, but it's a tiny place selling one thing: oil. There's a brief period in the 80s where the U.S. is growing over 8%, but that's coming out of a deep recession. After pandemics, after major wars, you can have temporarily high rates of growth. So if the U.S. were to screw up in some way or we had another pandemic, some kind of terrible war, some other weird event, and that combined with a burst of AI-driven growth that might have given us 3% to 4%. Could we for a year get to 10%? I'll say yes. I'd bet heavily against it, and the effect would be temporary, and a lot of it would come from getting out of the recession, war, or pandemic. But that's another way to think about it, fully consistent with what Alex was saying.
Erik Torenberg
And if robotics emerge, would you have a similar view? Because that is impacting more the real world, the real economy?
Tyler Cowen
Well, I think that would likely come relatively slowly. Look at how long it takes to get a Waymo in Washington, D.C. We're still not there. And that's a pretty simple kind of robot. There's forwards, backwards, and turns. And if you look at what are the valuations of robotic companies, they're way higher than they had been. Hard to estimate. But you don't just see them popping into trillion-dollar companies. I mean, you could count Tesla as a kind of robotics company, but the market data suggests it will happen. It's making good progress, but not going to give us 20% growth three or four years from now.
Erik Torenberg
Right. And so the smartest people who are the biggest believers in this runaway growth, the Drakash, the Leopold, our friends who believe that. What do you think they think differently? Is it that they have a different belief on the speed of technology? Is it that they believe that it will overcome some of these bottlenecks? Can you steelman the other side?
Alex Imas
I think, and Tyler, please chime in. I think the argument, I'm going to get a bit technical here. It's about having something called homothetic or non-homothetic preferences. Basically, what that means is that if you look at things that get automated in the past, what ends up happening is the thing that gets automated contributes less to growth. So let's say you have a million robots bringing robotics. The thing that they're making, because it becomes cheaper, ironically becomes a smaller part of the economy, and so they contribute less to growth. If you have non-homothetic preferences, which there was a really nice paper estimating these preferences using 200 years of data, what ends up happening is that, let's say you have broad automation, the value in the economy of what contributes to growth moves to the bottlenecks. That's just an inconvenient part of how growth is measured. And so the thing that's growing least will be contributing most to GDP. And that's why you get these sorts of effects.
Tyler Cowen
I agree with all that, but I would add a more sociological element. In my case, and I suspect Alex's, what is emotionally vivid in our minds is the long slog of human history, previous growth episodes, centuries of GDP data, productivity data. What I think is emotionally most salient, probably for Dworkesh, Leopold, and many others, is conversations they have in the Bay Area, which have real information. But when you ask the big question, with which emotional salience will you make better predictions, I would opt for what Alex and I have stuffed our heads with, or for what they imbibe all the time. But of course, they would take the other side of that.
Erik Torenberg
So basically what you're saying is when you look at the last few hundred years, the breakthrough technologies from electricity to the internet to everything in between, they haven't led to runaway growth. Or if they did, it was because of a small base or after a war or something like this.
Tyler Cowen
I do think this will likely be quicker than those, but still, you have a sense of the bottlenecks. Whether it's the Industrial Revolution, history of the Scientific Revolution, implementation of electricity, the IRS still uses fax machines, and so on and so on. Try using ed tech software, which I do all the time. It's a nightmare. They're not even up to how things should have been 10 years ago.
Alex Imas
And that's not regulation is the problem. It's a problem. But just inertia is difficult to overcome. And so much of our economy, it's government, it's nonprofits. It can be highly regulated parts of the healthcare sector, higher ed, general schooling. When you actually have experience with how those work, you could say 40% of the U.S. economy is a kind of nightmare. In Europe, it could be higher, like in France. And those sectors might need to be that way for other reasons. But still, it's a big dose of realism that not everything is like startups.
Erik Torenberg
Let's say you were both tasked by the president to come up with a set of metrics that determine whether AI is leading to our country's flourishing, broadly defined. Obviously, that's a very big question. Is GDP one of the main metrics you would think about? Should we be aspiring for it to have a big impact on GDP? Or what are some of the metrics you would consider?
Alex Imas
Yeah, this is a really important question. I think the other thing that I would add before even starting is, I don't think Tyler and I really disagree with the San Francisco consensus on the capabilities. A lot of people hear Tyler and I or people like us talk and they're like, oh, you're bearish on AGI or something like that. That's absolutely not it. It's more about how that affects the economy and the structure of the real economy. So thinking through what measures we can be using, I think GDP historically has been a really good tracker of human welfare and flourishing. It tracks pretty well with the gross domestic product.
But what Eric Brynjolfsson and Avi Kolos and others have been studying is the fact that technology has actually started to introduce this wedge where a lot of the products that technology creates are basically free. And so they think about digital entertainment and digital products. And then they have something called GDPB, or GDPD. I forgot the acronym. But it's essentially you take GDP and then you amend it with a survey based measure to actually see how much other products that are not necessarily priced or products that have become cheap or even free because of automation, how much the increased amount of those products is adding to human welfare.
Because you could see a world where you're still in this forward of 3% to 5% GDP range, which is already on the high side of growth. But you have a world that's actually truly flourishing. How do you pick that up? I think the way to pick it up is through these alternative measures.
Tyler Cowen
And longevity effects could well prove important. So, Eric, say you live to 120. I don't think I'm going to make it. That will in complicated ways boost GDP and in even more complicated ways possibly boost per capita GDP. But it is something distinct from GDP that you get all those extra years. And economics can value those years. And that's another way in which it may become more important to revise GDP. But as Alex said, GDP will become somewhat less accurate over time as AI progresses and will need to revise and improve it. But in the short run, it's just not that bad a measure. I like to say GDP, the most underrated statistic that there is.
Erik Torenberg
In the 2010s, Peter Thiel famously decried the lack of progress in atoms relative to bits. But here we are, 10 years, 15 years later, and we've got self-driving cars, early flying cars, we've got major companies that are now hardware companies, NVIDIA, SpaceX, Tesla. Tyler, would you go so far as to say that that isn't a concern as much anymore? Or how would you reflect on that 10 years later?
Tyler Cowen
The atoms will break through, I would say almost for sure, but we still don't have them. The contribution of driverless cars to actual human welfare is tiny. The biggest contribution is that it's fun and you can tell your friends you were in a Waymo. It's not really better than an Uber. For me, it's a little worse because they go by the speed limit. It's fun in a science fiction kind of way. Most of America still doesn't have it. Europe, forget about it. So we're still not there. That's the amazing thing. And it gets back to robotics also will be slow.
Erik Torenberg
I think I asked you 10 years ago, when would the great stagnation be over? And I think you said something along the lines of when we have self-driving cars. I can't remember if it was deployed or how diffused it was. But you still believe that?
Tyler Cowen
More or less. But that's a matter of degree. And they clearly work. And the new Tesla is pretty phenomenal. I've been in Alex Tabarrok's lately a few times. But in terms of market penetration, I don't know the numbers. But again, it's tiny.
Erik Torenberg
Let's go back to a different topic. Alex recently wrote a related paper about it, but I'll start with you, Tyler. The agent swarm topic, the Hugging Face incident, did that update your views in terms of what's possible or make you get concerned about something? Any reactions to what happened there?
Tyler Cowen
It didn't update my views much at all, and I'm curious to see if Alex agrees with this perspective. But as an economist, you think first and foremost in terms of principal agent theory. And unless the conditions are very restrictive, the principal just can't control the agent that well, no matter what contract they try to write. So when we replicate that result with AIs, well, it's interesting, but I'm not very shocked. The answer was always going to be checks and balances and what we might call separation of powers in the political realm, not you have perfect alignment. So again, if you come at this from the economist's point of view, all the hand-wringing about alignment, yes, people should work on that a lot. But at the end of the day, you're like, hey, I learned that when I was 17. It's a real problem.
Erik Torenberg
Alex, what's your thought on that?
Alex Imas
I unfortunately did not learn that when I was 17. I think we once again agree. There are these different perspectives on the issue of how do you best get AI to be in the service of humanity. That's the question of alignment. One way of thinking about it is to get the models themselves aligned with something like a constitution. The models at an individual level are aligned in this way because they have the same preferences as the human beings that are deploying them. But another perspective, which is what Tyler mentioned, is you align things, you don't even care what the preferences of these agents are.
They could be extremely individually misaligned, but you can create a mechanism around them where the incentives are such that the agent behaves in an aligned way. This is the mechanism design perspective in economics. As a principal, I don't need the agent to want to help me necessarily. The agent could really, really want to just help themselves and do whatever they want. And the whole study of mechanism design is how do you create the system of checks and balances in order for the agent to still be working in the principal's interest. From the research that I've been seeing coming out of the labs, both internally and externally, a lot of people are thinking in this direction that we really need to take a mechanism design perspective.
And as Tyler said, I wasn't shocked by what happened.
Tyler Cowen
I'm not aligned with the governor of Virginia, but it's a pretty well-run state, and I'm happy here. We have two-party competition, sort of, and it works.
Erik Torenberg
Tyler, do you think the way that Anthropic or OpenAI or some of these companies, or the AI safety community, thinks about alignment, do you think that is not the right way to think about it or confused in some sense?
Tyler Cowen
Well, there's how they think about it and how they talk about it. And how they talk about it is way too shaped by Twitter, which can be quite hysterical. How they actually think about it, I believe is a lot more reasonable than how it ever looks from the outside. So I don't have some big speech about, oh, we've got to totally reform them. I think they're pretty practical. They're companies, they have IPOs coming up. They don't want to get sued too much. They don't want to be over-regulated. We already have checks and balances. Not all of them good, to be clear. But the reality is the actual problems we've seen, the hacks, they're mostly coming or maybe entirely coming from evil human individuals where the AIs are doing what the humans told it to do. And that is a huge problem. But that is a kind of, well, we have the alignment. The evil human says, do this, and the AI does this. And that's where the costs are coming from.
Erik Torenberg
Some people make the comparison that alignment is in some ways like safety in general was in 2020, 2021. Around that time, there was more awareness around social issues. So CEOs had to say these things to please their employee base. Some of them were true believers and some of them weren't. And similarly, perhaps around AI safety, there's a lot of concern. And so some people are making that comparison that the CEOs aren't as true believers as they may be indicating.
Tyler Cowen
This checks and balances in these companies as well. It's partly why the companies work well, that you have to actually end up pushing through things that will work. And that's somewhat different than what everyone is saying. And there's a winnowing and selection set of mechanisms within the company. And people want promotions or bonuses, and they want to do things that can work. And again, it will end up going better than how it will sound on Twitter, I would say.
Alex Imas
And I think Hugging Face is a perfect example of, if you unpack what actually happened, it's just a bunch of human error on top of human error. It's great to have an incident. I mean, great in quotes, but it didn't cause a massive disaster. And it exposed a lot of broken incentives for the agents. So you see how the companies have reacted to the Hugging Face incident by learning from it and creating better environments. And so this is, nothing is smooth. It'll be a little bit down, a little bit up, but the slope is positive.
Erik Torenberg
Some people, Paul Christiano, I think, said in his speech, the most likely cause of death a year from now or two years from now will be an AI related death. And that to you sounds hysterical.
Tyler Cowen
Strong disagree.
Erik Torenberg
Alex, let's go deeper on your paper. Why don't you share some of the findings that you had and some of the questions that still remain?
Alex Imas
Yeah, so the paper, I believe you're talking about the organizing intelligence paper?
Yeah, with Andrew Coe and my colleague Shane Legg. The basic idea of that paper is that it starts with this puzzle: AI is becoming very capable on particular dimensions, but we don't see any visible economic transformation. The first thing we start with is that there's been this focus on individual agent benchmarks without thinking about what those agents are actually competing with in the economy. The relevant competitive unit is not the agent, it's the organization. The organization historically has been referred to as a form of super intelligence because of this concept called Ricardian comparative advantage. If you have a bunch of jobs, the person who does the job is not the average person, it's the person who's best suited for it. The envelope of the intelligence in an organization is way better than if you average across the population. These AI agents are basically competing against forms of human organizations or teams of humans. You shouldn't compare an individual level benchmark to how well a human does something, it should be across the entire kind of what we call mosaic intelligence.
Then we start layering on how capability improvements will actually affect the economy in equilibrium. This is a formal theoretical paper. We look at different levels of intelligence. We start with minimal AGI where the AI models are as good as average human capability in every single dimension. We show that even when you get to this point, you're not going to have a huge impact on the economy because organizations are not the average human ability, they're better than that. What you'll start seeing is wages change because production becomes cheaper, and that will affect human wages. What you actually need for real economic transformation is not just competence, like ability on particular tasks. It's the ability of these agents to coordinate with one another.
If you look at Hugging Face and even the solutions to the millennium problems, there's this idea of agentic swarms. But if you unpack it, these agents are still very bad at coordinating. They're not coordinating, they're mostly just doing parallel processing. If you have benchmarks specifically on coordination, you can show that one agent does better than two agents splitting the work, even if they can communicate. Whereas if you give the same thing to two people, they actually have division of labor, they communicate, and they do way better than one person with double the resources. The agents are still pretty bad at coordinating, but if they get better at coordinating, you can break the cohesion ceiling on firm size and get very different firms to emerge. That's the basic premise: once coordination becomes better, the economic space of possibilities of what firms look like, what happens to wages, what happens to employment, that's when the real visible changes are going to happen.
Erik Torenberg
What's the biggest bottleneck to figuring out coordination at the moment?
Alex Imas
To be honest, I don't know. I think it could be as simple as it's not being trained into the models. It's not part of post-training. I actually don't have any insight on that.
Erik Torenberg
We've talked about some areas where you agree. I'm trying to find some areas where there's some difference of opinion. Maybe let's try employment. Alex, I think you're more worried about the transition happening so fast that there might be something sudden happening on the employment side. Why don't you share your concerns or views?
Alex Imas
I'm basically worried about the potential. I actually agree quite a bit with Tyler that things are potentially going to move pretty slowly as far as the effect on the real economy. But if there is a fast transition, let's say there's a breakthrough in coordination, what we show in that paper is that there is the potential for wages and employment to suddenly have a phase transition as far as what will happen to the economy. In that case, there still will be new sectors and new potential opportunities. But the rematching that we have in the economy that used to take quite a long time doesn't have an opportunity to happen. So you have larger pools of unemployment, and that leads to recursive loops where people get scarred out of the labor market. That's my personal worry. As far as likelihood, that's a different story.
Tyler Cowen
I have a related worry. It's close to the same thing, but it's just hard to tell right now. If you're hiring someone, will the person be good at AI? The skills they used to need, we had imperfect indicators. They went to Cornell, or they got all A's, or they had a good interview. Now the skill is, will they be mentally flexible enough to pick up AI? We don't know often. So that could increase the natural rate of unemployment. Good matches will be harder. Search will take longer. I don't think that problem will be with us forever. Eventually, the AIs themselves will figure it out, and there'll be certifications for how good a person is at AI. But in the transition, everyone's a novice. You might be well-established the way a lot of people in the tech world are, but otherwise we're all novices at this very important new skill. So I guess I'm curious how you think about employment, wages, these questions.
Erik Torenberg
Well, some additional slowness in hiring people. But there'll be an incredible number of new projects, new jobs. So I'm pretty bullish, actually. Labor markets have been fine. But of course, there'll be disruptions and transition problems, as there always are, as there were with the China shock, the Industrial Revolution. So I expect a new set of those. A bunch of them we haven't even figured out or seen yet, but they'll come. Absolutely.
Tyler Cowen
To what extent will there be just an explosion of agents buying and transacting with other agents in this kind of almost parallel economy? There are so many more of them. They're 24/7. How do you think about agents becoming the dominant force on the internet? What is the right mental model of thinking about what that will look like or what that won't look like?
Alex Imas
I don't know what you mean by the word explosion. If you ask me when will agents be 10% of GDP, I don't think it's soon. If they rose from basically zero to 6%, you could call that an explosion. 6% of GDP is enormous. Even that doesn't seem that soon. If it's 1.5%, I could see that within five years. If you're going to force me to give guesses, you could call one and a half percent an explosion. GDP is huge. That's one of the core insights of economics. GDP is huge.
Tyler Cowen
Even with the agents that we have now, they're already incredibly capable, like the consumer grade things that live on your phone. But there are completely unintuitive bottlenecks out there. Websites are designed for human beings and they're all different. So you're still getting into issues where your super smart agent is booking you a flight, booking you a seat right next to the toilet every single time. Then you're like, I don't want to use an agent next time. Or the fact that there's all of this regulation out there. That's the big one. Companies are spending billions of dollars trying to do basic choice architecture on their website in order to upsell humans on products. Agents don't have the same psychology, so they're not going to be prone to this choice architecture. These platforms have an incentive not to allow somebody else's agent to go in and shop for them. You're already seeing it. If you pull up Muse or something like that, it can't shop on Amazon for you.
I don't use agents to manage my calendar, although they could, because I want to know mentally in my mind what's coming each day. The only way I can remember is if I have to write it down in extra time on my calendar. That's stupid, that's my flaw. But I don't think it's going away. The agents could be 10 times better, whatever that means, and I'm still going to want to write it on my calendar. I think, oh, December 18th, I'm seeing a Renaissance music concert. I know that off the top of my head. There are these funny indivisibilities in how we allocate our attention that are often some of the biggest bottlenecks.
Erik Torenberg
Going back to the regulation part, Tyler, you seem more sober that the midterms might not lead to data centers stalling indefinitely. I'm curious for your broader thoughts on AI sentiment. How much of the negativity is real in the sense that it will lead to things like Bloomington trying to outlaw usage? What are your thoughts on the political economy there?
Tyler Cowen
We have 50 states. There's counties, there's cities, there's even other countries. It seems it will slow down a bit, but still basically go through. The new Texas governor is already backtracking. They're like, well, we just want the labor unions involved, they need to pay good wages. The projects we have now will go ahead. The others we're going to look at again. Josh Shapiro, the rhetoric was very strong. The reality is quite different. So it raises costs, but I don't think ultimately it will matter that much. Will there be a bunch of places that absolutely say no? Probably, but again, 50 states. I'm glad we have them. Maybe we'll end up with 52.
Alex Imas
I think it completely depends on the path. Everything is kind of endogenous, including political economy. The reason the governors are backtracking is because the unemployment rate looks great, the economy is growing at a reasonable amount, everything kind of looks fine economically. There's no instrumental reason for there to be this political economy reaction. The thing that people who are projecting these very extreme scenarios don't realize is that with the beginning of an extreme scenario, there will be the beginning of an extreme political economy reaction. That will endogenously constrain things. It's just going to be a back and forth. What the data will end up looking like to people living in the future is that things moved a lot slower than people would think because of this interplay.
Erik Torenberg
It seems like the amount of money one can put in politics, everyone was worried about Citizens United a long time ago, that corporations are going to control everything. But Elon put in 500 million or whatever it was and couldn't get the outcome he wanted. Is this an exposed limitation of money in politics, or is it just that tech people, or not just tech people, but corporations may be very good at allocating for their business, but maybe not that great at allocating in the D.C. and political realm?
Tyler Cowen
Dare I say that academia is slightly underrated by Bay Area people? People who study campaign finance have known this for a long time. There's a lot of evidence that spending doesn't matter that much. You can think that's good or bad, but none of this surprised me in the slightest. I don't think it's that the people are fools. You might think they're fools in some broader sense of just having political aspirations at all. But there's just no simple way to spend a lot of money and get the politics you want. It's checks and balances. Even though no one in the system is aligned with you. It's a beautiful system we have.
Alex Imas
I'm not joking. The American education system is incredible and competitive markets are incredible. It's just everything, it's not perfect, but exactly as Tyler said, there's all this magic that comes out of the system.
Erik Torenberg
I'm curious if you were both interviewing each other, what you would ask. What are you most curious about? Tyler, I'll start with you. Where are you most curious about Alex's perspective on something?
Tyler Cowen
Well, your origin story of how you became an economist?
Alex Imas
That's a very self-serving question, Tyler, because it has to do with you.
Tyler Cowen
Okay, we'll say more.
Alex Imas
I'm an immigrant from Moldova. I grew up on food stamps and things like that in America. My parents didn't know anything about economics. I didn't know anything about economics. I went to college and I was pre-med because I knew what a doctor was. My parents knew what a doctor was. In college, I had to take prerequisites. I took an econ one class and I thought this was kind of fun, but I'm still pre-med. I got good grades, made my GPA higher than my pre-med classes suggested. I was going to go to medical school. I applied, I got in. Then I started reading Marginal Revolution and I started listening to an NPR program with Richard Thaler. Between the both of you guys, it was just, oh, I can do this with economics. This is incredible. I only knew one part of that story, so thank you. I knew the revolution part.
Tyler Cowen
No, I mean it was because you were at that point, this was 2008 or 2009, there weren't a lot of economists showing you how much you can do with the topic. So I pulled out of the medical school process. I went to the only school that accepted me, which was UC San Diego. The rest is history. This is just such an incredible field and knowledge base to understand, to the extent that we can, which is not very high, about what's going on in this complex system that we call the economy.
Erik Torenberg
Can I ask a follow-up? Will and should Moldova become a part of Romania?
Alex Imas
Yes, that's my view. And they will. I figure they should, but they will.
Tyler Cowen
They were originally. The country was called Basarabia. My grandma, she was from Basarabia, she wasn't from Moldova. This is probably a controversial opinion, but I'm actually not from Moldova. I'm from Transnistria.
Ah, okay. Now it's all clear.
Alex Imas
We were refugees from Transnistria to America. There was a civil war. We were in basements hiding from bullets and all that type of stuff. Then we ended up coming to America as refugees. If you know Moldovan culture, it's very similar to Romanian culture. The language is almost the same.
Erik Torenberg
When the VCs say always ask for the origin story, they're so correct.
Alex Imas
What question would you ask Tyler?
Erik Torenberg
I want to know, because you were so early on this and I don't think I've ever heard this story of how you were attracted to AI and thinking about AI so seriously. You were very early about the transformative potential of this technology. I'm just curious how you realized that earlier than most.
Tyler Cowen
All chess players have a story like this. I was in early years a chess prodigy. I was age 13 playing in New Jersey, and they wheeled in a computer called Tinkerbell that played against the humans, and it sucked. But I saw what the progress was like. The point about chess players is this: they all know if they're any good. Chess is not like pocket calculator stuff. It's intuition. If you see computers conquer an area that you already knew was mostly intuition, you will find it very easy to believe they can conquer a bunch of other areas. Kasparov was early to this. Ken Rogoff didn't need much AI pilling. He was early to this. The chess people were the first ones there. That's a very simple, direct story. That's how it happened. Demis too, absolutely. Common stories. You see it kick butt by the 1990s and it beats Kasparov. Deep Blue is not a great model, but it can do this. So you're just waiting for it to spread to everything else. But it took a long time, in fact.
Alex Imas
I think this is a great example. A lot of people would have looked at AI and said, hey, it can't do this, it can't do that. But it's going to be able to do almost everything because thinking about RLVR or something like that, the verifiable reward part of post-training, a lot of people use that as copium. Oh, that's not verifiable, that's not verifiable. But if you train the verifier, almost anything will become verifiable. The fact that you can criticize it means it's verifiable almost.
Tyler Cowen
Exactly. Chess is a really great example. If you're a really good chess player, you know exactly how much intuition is required and how far that's going to go in the future.
Erik Torenberg
Mercore is working on the poets. I'm not sure how far along they are, but it seems to me it should work. Create better poetry. To write great poetry. I think in less than two years, it will be possible if people put their minds to it at all. I don't mean a big major effort with billions, but just some kind of effort at all. Bring in a dozen poets, pay them something. Doesn't have to be that much. Let them loose. Have some checks and balances on the process. It'll work.
Alex Imas
Some people at the firm think that the idea of taste being uniquely human is a joke. The systems are already good at it. They already have great taste. You ask them about classical music, what's the best recording of Sibelius's fifth, whatever. They give great answers. Better than the Penguin Guide right now.
Tyler Cowen
Fascinating.
Erik Torenberg
Okay, there's a few grab-bag questions I want to ask before we wrap. One is, did you guys see Dorkesh's essay he collaborated with somebody on, basically saying that Piketty wasn't right about the last century, but his model of RRG might be right in the AI century? I'm curious if you had a take on that, Alex.
Alex Imas
This is with Phil Trammell, I think. Capital in the 21st century. The idea is that economic activity may be divorced from labor and be kind of focused on capital. Many people had issues with the actual methodology of the piece, which I won't speak to. But the broad idea of the technology potentially increasing the reliance on capital, that is a scenario that could happen. There's alternative scenarios, things that I've written about and Tyler also alluded to, where there's an alternative scenario where there's just more jobs created and these jobs are valuable because they're scarce. Value moves towards scarcity. That's the bottleneck logic. If those jobs are in a rich society, there will be a lot of value going to those jobs. It's just a question of how many of those jobs exist and whether it can support the entire economy. Historically, that's been the case, but that's the real big question mark. These are things we just need to watch for with data. My paper on economic policy for AGI makes that argument. We need to be tracking where the economy is going with AI in a much more comprehensive way.
Tyler Cowen
Very fundamental technologies tend to benefit the consumers. Fire, language, electricity, and so on. That's no proof it will happen again. But we need to distinguish between capital can do a lot versus there are entry barriers that will accrue monopoly rents to particular capital owners. I don't see that piece really made a strong case for the latter proposition. It was mainly arguing capital can do a lot. But if the capital is commoditized, I know people who say all the rents will go to land, all the rents will go to energy. I'm not sure that's true, but those seem more likely than all the rents go to capital.
Alex Imas
That's right. It goes to scarcity. If scarcity is energy, that's where the rents will go. Now, maybe AI itself makes energy less scarce. So that's why I'm reluctant to predict. I at least see the case to think the Gulf countries will make out like bandits and so on, or the landowners. But AI will change enough things. Say we can terraform land quite cheaply and desalinate water very cheaply. It doesn't seem so implausible. Maybe it's location of land or the best retirement spots. Lake Tahoe is so special because you're near the Bay Area but you don't pay the taxes. It's very hard to predict. But think through scarcity most of all.
Erik Torenberg
In our world, we think the big scarcity is on supply. In the short term, demand is much bigger than that. But that rests on the belief that when you talk to the super users, they can't get enough. The super users are really a small percentage of people, a small percentage of companies. AI is still not broadly diffused yet. The counter to that claim would be, well, it's not having measurable impact on companies yet. So it is speculative that the more AI is used, it will eventually lead to impact on these companies in ways that justify the massive increase in spend. We just started a new fund called the Machine Age Fund, which is doing data centers and other elements of the infrastructure stack to mitigate some of these supply challenges.
Tyler Cowen
Leopold's paper, Situational Awareness, does that stand to you as one of the more important public intellectual pieces in technology in the past few years? I'm curious what you think of the legacy of that.
Alex Imas
In anything, not just technology. But that said, I think he's too enamored of fast takeoff. As I told him at the time, he's assuming simulation can substitute for experiment. Thus you get this never-ending loop where science just keeps on accelerating. I don't think simulation very much subs in for experiment. It will in some areas.
Tyler Cowen
Alex?
Alex Imas
We agree. I'm too Lucas critique-pilled for that. Sorry, inside baseball. But it's basically why you need to understand the deep structural parameters of the problem in order to actually be able to predict. What simulations do and what AI is really good at is predicting, but it doesn't take the endogenous response of the system into account a lot of the time. We'll see that play out more and more, I think.
Erik Torenberg
Robin Hanson's Age of M. How do you think that stands up or what would the legacy of those ideas be?
Tyler Cowen
I think it's wrong. It's a brilliant book, and I love Robin. But when it came out, which is now a while ago, I said, Robin, AI is going to get there first. Robin says, no, AI might get there, but M has to come first. I say, Robin, M probably will never happen. But if it does, first it needs AI. I think my side of that debate is looking very good. But it's a great book to read for getting ideas and just shaking you out of the dogmatic slumbers.
Erik Torenberg
We're talking about NVIDIA being in a big crosshairs at the moment with the Jensen situation. As you view the public discourse, what do you think is misunderstood or how are you interpreting what this potentially means? What lessons to take?
Tyler Cowen
We've already said alignment is impossible. The stakes were so high. His behavior, the company, something seems out of alignment. In fact, it will be fine because of checks and balances. I don't know how they're going to operate. I think it's a great lesson in all the different things we've been saying.
Alex Imas
I think this is not really going to change the trajectory much. Checks and balances work. We have diversity in the supply chain to a large extent. Everything is over-subscribed. Demand creates supply. Supply creates demand. The system is robust. The biggest change might be if something goes wrong at the macro level and NVIDIA had to step in as a buyer or lender of last resort with a weakened or gone CEO founder type. They're a lot less able to do that than they would have been had this never happened. That matters. But in the medium term, it's going to wash out. It could slow things down a bit, but the structural forces will take over sooner or later.
Erik Torenberg
If you had to expect what is the next genre of situational awareness-like document that is going to be remarkably prescient and greatly impact our society, where would you expect that to be? Would it be in something in biotech? Would it still be in something in AI? How are you thinking about what are some underrated ideas right now that might be massive in the years to go?
Tyler Cowen
They're underrated to me too. Everyone talks about biotech and I'm fully on board, but I think that will be especially slow because of regulation and the inability to use simulation to address the main issues. But again, Eric, I think you'll live to 120 and I think I will not. That's the simple way to put it.
Alex Imas
I think we're going to get better models of micro and macro phenomenon, better structural models. Something like the model of the cell or using digital twins to do macroeconomics. To the extent that you are going to be able to put structure on things that we need to use to predict, that's going to be the real big blow up. That gets to that idea in economics, the reduced form versus structural debate, where you understand the structure, you can really create brand new things. The potential for AI to help you understand and recreate that structure, I think that's really underrated.
Tyler Cowen
It'll be great. Ross Chetty has a new paper out on digital twins.
Alex Imas
Exactly, digital twins. That's exactly what I'm referencing. But he needed five co-authors and a lot of money and a research center, which is fine. It's wonderful that he's done it. But that also goes to show you something.
Erik Torenberg
Let's close on the university question. We said earlier the academy is underrated in some ways, certainly by Silicon Valley. Now that we've given the props, I do want to ask you guys, where are you most excited in terms of innovation in higher education? Where is the most urgent need? We are interested in the question having incubated Horowitz and Dresden Academy. Maybe Tyler, we'll start with you.
Tyler Cowen
I hope you all do something and spend more money on it than what you've announced and put some of your best people on it. I don't really know what your plans are, but I'd love to be involved in some way. As with you, Austin, and Minerva and other innovations, we need much more experimentation, innovation, and measurement in higher ed. If you're a part of that, I'm thrilled.
Erik Torenberg
Amazing. I would love to have you. Alex, where are you most excited or what advice might you have to anybody trying to innovate in higher education?
Alex Imas
More innovation, the better. Higher ed is going to change drastically in the next, I don't have a great timeline for it just because they're such slow-moving institutions. The forces of the market are going to lead to a shrinking of our standard idea of higher ed, the traditional version. But we're going to have a ton of new innovation. I'm very bullish on higher ed in the long term.
Erik Torenberg
Are you both bullish on the field of economics at the moment? Are you encouraged by what's happening right now? Where do you stand there?
I mean, I think the economic toolkit and the economic way of thinking is extremely underrated. So I'm extremely biased too. But I think the way that I see things play out, we had the earlier question about alignment and the economic toolkit is perfect for thinking about that through this mechanism design perspective. And you have this type of thing going on all the time about endogeneity and the political economy question and thinking through these sorts of consequences, thinking through the Lucas critique about how you estimate a parameter, but that's actually not a structural parameter. So it's not going to be valid to actually use that parameter in order to make the decisions. So all of these sorts of things are super important and they're going to be more important in the coming days, I think. So I'm bullish.
Tyler Cowen
And it may just become social science. So the things economists don't know, the AI fills in for, the basic mental toolkit becomes more valuable. This is originally a Chicago idea from Gary Becker and others. And you just want really smart people who have some skills a bit like those that we now call software engineers, but it will have its own name. And they have their fingers on the dials, and they manage a bunch of AI agents, and they're empresarios, and they know economics and do all of social science. That's what I expect. So a true realization of the earlier Chicago vision. I'm from the University of Chicago, so I buy that vision. I'm happy with that vision.
Erik Torenberg
Maybe we'll do it at the University of Chicago. Amazing. Full circle moment to end the conversation. Tyler, Alex, thank you so much for coming on the podcast.
Tyler Cowen
Thank you.
Alex Imas
My pleasure.
Erik Torenberg
Thanks for listening to this episode of the a16z Podcast. If you liked this episode, be sure to like, comment, subscribe, leave us a rating or review, and share it with your friends and family. For more episodes, go to YouTube, Apple Podcasts, and Spotify. Follow us on X at a16z and subscribe to our Substack at a16z.substack.com. Thanks again for listening, and I'll see you in the next episode. As a reminder, the content here is for informational purposes only. It should not be taken as legal business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any a16z fund. Please note that a16z and its affiliates may also maintain investments in the companies discussed in this podcast. For more details, including a link to our investments, please see a16z.com forward slash disclosures.