Odd Lots - David Shor and Byrne Hobart on the Politics of a White-Collar Wipeout

Episode Date: March 24, 2026

Nobody knows when or if AI will lead to mass displacement of white-collar work. But the anxiety is clearly here now, and there's very little evidence that our politicians are taking it seriously. Of course, there are at least two questions operating at once here. The first is whether or not AI really poses a significant threat to the existing labor market. And then the second one is about the correct policy response. This was the subject of a recent Odd Lots episode recorded live at SXSW in Austin, Texas. In this conversation, we were joined by David Shor, a political consultant, pollster and founder of Blue Rose Research, as well as Byrne Hobart, the writer of TheDiff newsletter, and a general partner at Anomaly Fund, an early-stage venture capital firm. We discuss the prospects of a labor market disaster, what David's polling says about the public view, and possible policy considerations that could be palatable to both industry and the general public. Read more:Fink Says AI Threatens to Leave Masses Behind Unless They InvestPrivate Capital Turns to Old Economy as Software Trade Dims Only http://Bloomberg.com subscribers can get the Odd Lots newsletter in their inbox each week, plus unlimited access to the site and app. Subscribe at  bloomberg.com/subscriptions/oddlots Subscribe to the Odd Lots NewsletterJoin the conversation: discord.gg/oddlotsSee omnystudio.com/listener for privacy information.

Thanks for listening to OddLots. Follow the show on Amazon Music for more future episodes or just ask Alexa play the podcast OddLots on Amazon Music. Bloomberg Audio Studios Podcasts Radio News. Hello, OddLod's listeners. I'm Joe Wisenthall. And I'm Tracy Allaway. Tracy, we were down at South by Southwest recently. Yeah, that's right. I didn't get to have any barbecue in Austin, but we did have some really good Mexican food. We did have really good Mexican food. And we recorded a really restaurant. really fun episode of our podcast. That's right.

More importantly, we recorded a good episode. That's right. So check it out, listeners. We did a live episode of the podcast talking about sort of the prospect of mass white collar displacement due to AI as well as the politics and how politicians should be thinking about this and how voters are thinking about this. Our guests were David Shore, founder of Blue Rose Research and Bern Hobart, who writes the excellent diff newsletter and a general partner at Anomily Fund.

Take a listen. Thanks, everyone for coming out on a cold Sunday morning. Yeah, hello and welcome to a live recording of the Oddlots podcast. This is definitely going to be the most uplifting of all the conversations that have happened this weekend, right? I think so. I think it really will be. Everyone here is going to leave feeling really good about the future.

All right. Well, let's kick it off. So we have two great guests. We are going to be speaking with David Shore. He is the founder of Blue Rose Research, Political Consultancy, Polster. knows a lot about AI, and we have Byrne Hobart, the founder of the diff, a great newsletter that everyone should read and a general partner at anomaly funds. And we're going to talk about all things, AI and job loss potential and the politics of it and so forth. And so David and

Byrne, thank you so much for joining us on stage here. Great to be here. Yeah, thank you. So let's just start on this question. David, you're pretty like, this is happening now. the economy is going to look radically different, maybe even a year, 18 months from now. And you're trying to wake politicians up. This is happening right now. But like, tell us what's happening right now or what's about to happen. Yeah. You know, I'm not an AI expert, and I don't want to claim I'm one.

But, you know, what I'll say is that I use AI a lot. I use cloud code a lot. I think in December I spent maybe like 15% of my pre-tax income on cloud code overage fees. And I think there's a real disconnect among my friends, between, you know, people who use cloud code and people who don't, where I could really feel the extent to which I can do so much more now versus a month ago. And a month before that, like, the scale at which these things are getting better and the speed at which things are getting better is really jarring. And I think you might not notice that if you're just using chat GPT.

Like, you know, chat GPT is like a little better than it was a year ago. But again, it's very hard to make predictions about the future. But I do think that if things continue to improve on this scale, then things could get really weird really quickly. And, Bern, one of the reasons we wanted to talk to you is because you're sort of at the intersection of technology and finance. If I look at the valuations of big tech at the moment and a bunch of the AI companies, they basically suggest that they're going to take over the entire economy. When you see those numbers, what do they suggest to you about the future of labor? So I think right now it's definitely true that you can think of this coherent category of this is an AI company or this was a non AI company, but they're incorporating AI and they have these distribution advantages.

So they'll probably do well at that. But I think that it's a mistake to think that this is a durable, discrete category in the same way that you could refer to a lot of companies as electricity companies. And if it were 1925 and you're trying to figure out which stocks to buy and you're just really, really bullish on electricity, then you would really, really care that our. CA is exposed to electricity, and I don't know, U.S. Steel is mostly not. But eventually every company becomes an electricity company, just in the sense of you would run a very different business if the lights did not turn on. And so it like gets subsumed by the rest of the economy. And you see that with software too, where there are just a lot more companies that have software developers and are writing software for internal use. And they're not software companies per se, you know,

their restaurants or like tractor companies or whatever. But they have that element. And so I think That's part of what you see just early in the role out of any general purpose technology is that you have a lot of really narrow specific bets. And then over time, the impact gets so widely distributed that it's very hard to trace. And you have to kind of go back and look at the history and look at things like, okay, the rise of the car leads to the rise of the suburb. But it also leads to the rise at the grocery store because, or like the supermarket where you can have a much larger selection and that means lower labor costs per unit sold. And that means lower costs overall. And that works if people are not walking to get their groceries. It doesn't work if people are walking and it's like a daily, you know, stop on the way home from work or something.

So we will probably see a lot of those weird kinds of outcomes. Like I think if you're thinking about the risk to a given career, I think for most of the careers people worry about, the average, like the mean compensation goes up, the median compensation of people who are in that industry right now. The median compensation they get from being in that industry probably goes down where a lot of people get washed out. But this has happened before. Like the spreadsheet did not eliminate investment banker or accountant as a job. It actually made it a more lucrative. job, but it also made it a more measurable one where you just can't slack off the way that you perhaps used to be able to somewhat slack off in some of the white collar professions.

It's just a lot easier. Like the expectation for output is so high. I'm on Twitter all day and a lot of people are tweeting and I'm like, it seems like a lot of people are slacking off these days. I'm like, how do you have time? I have time because I'm like a journalist and I said, you know, I create words professionally. Some would argue that you do not in fact have time to be tweeting all day. That's true.

But David, so, okay, there's been a lot of progress. No question. There are new harnesses for AI models that increase all of our capabilities and so forth. That's different than, we all know that's true. That's different than like job wipeout in a significant way. What is it about to you that you think, yes, there is progress, but progress on the scale of this is an imminent thing that we have to be talking about that could really reshape white-collar labor. The analogy that I think about a lot is COVID, because the thing about

AI progress is just that it's really in many ways exponential, where the amount of time that an AI can operate autonomously without a human really has been doubling every 112 days or so for the past six years. And, you know, when I think about COVID, this was a thing that nobody saw coming. and then it happened really fast. And then I think the political system was very reactive. And I think that a lot of quietly, a lot of Democrats wish that they had handled things a little bit differently. But what's cool is that unlike COVID, you know, we really can see this coming.

All of the warning signs are blinking. And, you know, the other point I want to make here is, you know, I personally think that there's a lot of potential for large-scale job loss, particularly white collar. But, you know, there are a lot of truck drivers. There are a lot of Uber drivers. You know, all kinds of people could lose their jobs. And this could all happen very quickly. I think the more important point, though, is that the American people see this and are really quite worried. You know, when you ask people, how likely do you think it is that in the next five years, there might be large-scale job loss because of AI, 70% of the population says that it's either very likely or somewhat likely. And so that's really the main point I'd make there is the reason politically why politicians should act now. rather than waiting until there's a problem is, you know, one, the American people are already worried about this. And two, once it's already happening, it will be too late for our political

system to respond. And so I think it's important to try to get ahead of. Definitely want to get more into the politics. But, Bern, you brought up something that seems to be kind of standard in these conversations, which is we've been here before, right? We all, well, not literally, went through the industrial revolution, but that's something that happened. and we went through the internet boom, and the economy and society adapted, more or less. But when you ask people now what the alternative professions are for white-collar workers, we haven't been able to really get like a slate of possibilities.

I know it's hard to imagine the future, but, you know, if you're an insurance broker and you have been for 20 years, what are you going to be doing in the new economy? Like, what are the new jobs that are coming down the line? Yeah, that is actually a tough question. And one of them is just temporarily, I think there are a lot of jobs that are basically either human who's required to be in the loop for regulatory reasons. Like, doctors are incredibly rapid adopters of AI tools.

And the supply of doctors is sort of artificially constrained. And so if you decrease the percentage of their time that they spend on admin tasks and you decrease the rate of mistakes that they make, you basically get the equivalent of manufacturing more doctors. And in cases like healthcare, there is effectively unlimited demand. Like there's yet to be an economy where people don't spend more of their marginal dollar on health. So we're all going to be health care workers. I think that's actually like that sector probably will grow.

And, you know, it's kind of glum to say, okay, some some white collar workers are going to kind of move downscale in terms of status where they will probably have jobs that sound less cool. But if overall output is high enough and if the returns on capital are high enough that people like the economy starts shifting more incremental production into just building data centers, that does mean that. It does mean that if you are the complement to a data center, like you are a necessary component of this entire supply chain, your bargaining power is a lot stronger because there's just a lot more capital that you're adjacent to. If you are completely substituted by the data centers, then you're in a tough spot.

But we always find that models have these really spiky abilities. Like they're superhuman in some respects. And in other respects, they are, you know, in terms of things like math ability, like they're beyond the point where I could reliably distinguish between two models and say, well, this one's really good at math this is okay at math. But in other domains, they do just kind of fall flat because they don't have this comprehensive world model. And the reason for that is that they're trained on text,

and text actually skews towards areas that are uncertain and open to debate. So I use the term the maybe sphere, which is like, if you imagine there's like this bedrock of facts that are so obvious that nobody ever bothers to write them down. And then there's this infinite space of questions that are so weird that it's almost certainly don't have an answer. There's like this little narrow layer like an atmosphere where it's worth asking a question and you might get an answer. So they have a really good world model for the parts of the world that we're either not sure about or that we've codified in textbooks and then a really bad world model for a lot of the obvious stuff. And so it would be kind of weird to be like to say like my job right now is to say extremely obvious things to the superhuman

intelligence. I don't even know what what job title historically that might correspond to. Like sort of a servant for, you know, a brilliant person who's also like a, you know, absent, Minded Professor. But yeah, I think we'll have a lot more man-servant for absent-minded professor in the future. That's future. Man-servant for absent-minded professor. Sounds fine. Canadian women are looking for more. More out of themselves, their businesses, their elected leaders, and the world are out of them. And that's why we're thrilled to introduce the Honest Talk podcast. I'm Jennifer Stewart. And I'm Catherine Clark. And in this podcast, we interview Canada's most inspiring women. Entrepreneurs, artists, athletes, politicians, and newsmakers.

all at different stages of their journey. So if you're looking to connect, then we hope you'll join us. Listen to the Honest Talk podcast on IHeartRadio or wherever you listen to your podcasts. David, you have a firm and you hire people. Has the nature of your hiring changed? Are you hiring for different types of roles

than you would have a few years ago or something like that given the change in technology? Absolutely. You know, I think just to give a, simple example, like we used to have lots of copy editors to write polling questions or to write messages. And, you know, the reality is that now AIs are generally better than people at doing that, not uniformly, but we have a lot. You know, they good at writing polling. Like, can AI, you want to find interesting polling

questions, right? Not the obvious stuff. Is AI good at finding non-obvious polling questions that are non-correlated to think so that you can actually get signal from them? Well, you know, I don't want to talk too specifically about that. But I, but I, I will say that there's just tons of copy editing tasks that I think, frankly... I mean, why do you want to talk specific about? Yeah, yeah, I know. That means this is the question I should be asking. No, no, no, no.

But, you know, look, there's a lot of stuff like translation. There's a lot of stuff like copy editing. I think the big shift for us is that we're focusing a lot more on person-centric jobs. And, you know, now you can do a lot more engineering than before. So I think there's definitely been a big job shift, you know, in terms of how we've been hiring. Absolutely. Maybe we can do a little bit of content creation, naval gazing here. Please.

I imagine this is probably of interest to a lot of people at South by Southwest as well. But, Byrne, you write a newsletter on a daily basis. Joe and I do as well. How are you using AI just in your sort of day to day? So I use it a ton for research. And one of the specific use cases is asking a questions like, does this joke land, or, hey, I'm making a statement, you know, kind of narrow technical statement about a domain that I'm reasonably familiar with, but I'm not an excellent.

expert on, you are an expert on this, tell me what I'm getting wrong. And one of the reasons for that is just a lot of my readers are software engineers and are eager to tell you when you're wrong. Yes, extremely eager. Like I sort of measure people's ability as software engineers in particular based on how quickly I get an email from them that there's a typo. Because the really good ones, like one of the skills that they have is just looking at a lot of text and immediately seeing what's wrong. Now that skill is kind of obsolete right now. LLMs are better at it. But still, the mindset of I'm going to look at this and kind of absorb something coherent and then I'm going to look for any little issue with it, that's still quite valuable. In fact, since there's more code being produced

than ever before by a huge margin, it's really, really valuable to be able to read it fluently and understand what it's supposed to be doing. So I do, I use AI a lot for research. I have only in the last few months have I actually gotten ideas for things to write, specifically from chat GPT, or only in the last few months have it made some original points I would not have thought of that were just like clever insights and not just it's reciting a fact that I did not have happen to know. David, talk to us a little bit more. So this question of progress, right? We all know it's getting better. That's obvious. But the more interesting question is, is it getting better at a pace faster than what people had previously anticipated? And talk to us about like, okay, where we are

now with the technology and where the people that you were talking to, and maybe I'll throw this to both of you, where would they have said we would be in, was it March? March 2025. Yeah, I think that the big surprise of the last year has been, you know, the rise of vibe coding, the rise of tools like Claude Code. If you just went back a year ago, I think that nobody really expected the extent to which these things would be able to do large-scale, complex, autonomous coding problems. You know, I think in a lot of ways, these models have become useful faster than they've become smarter. And I don't think that that's something that people expected. But what's interesting is that every year there are a bunch of AI experts who then go and make predictions. And one of the biggest surprises has been the revenue growth of these companies, which I think is in many ways the most important benchmark.

I think Anthropics revenue last year was something like two X, what experts who already, I think, were quite AI-pilled predicted. And so I think that's probably the biggest surprise. And, you know, what I would say just on the broader transition questions is just, that tools like LLMs have been now adopted faster. You know, there are a bunch of graphs that are like, how long did it take to implement radio or electricity or the internet? And this is much faster than any of those things. And so, you know, I really worry, like when we talk about past transitions of like

telephone switch operators or factory workers, you know, all of these things happened over an extended period of time and only impacted certain sectors of the economy. And this technology really threatens to upend every single job at the same time, you know, at a point when, you know, people's overall views of the economy are not good. And so I really worry just politically the extent to which our political system can handle this. It is true that just when you look at these measures of how broadly technology is adopted, I think the thing that's really hard to measure is at what point does it become just part of your baseline expectation and at what point are the impacts so obvious? that they're unspoken. So if you look at electrification, like, I think anyone who talks about AI is just required to say that it took like half a century to go from the first electrified factory to most U.S. factories being electrified. And the reason for that, and there's a lot of fun

economic history on this, is that you have to run your factory in a different way. You actually have to build a different kind of building. And you also have to finance it a different kind of way, or you can finance it a different kind of way. So if you have a factory that has some kind of mechanical power source, you tend to expand your business in these discrete factory size. increments. And what that means is that to your investors, you pay out most of your earnings as dividends because there's nothing to do with retained earnings. Like if you're going to grow, you want to issue a bunch of stock, you want to issue a bunch of bonds, and then grow in like one shot. But once you have electrified factories where they can expand, they can just add another assembly line or they can

upgrade this machine to a newer machine, et cetera, they can actually expand more organically in incrementally. And so that's when you start to see dividend payout ratios come down. And that's when you start to see the growth company as a concept emerge. Like if you read investors, accounts from the 1920s and they're talking about the stock market, one of the weird things is they're very fixated on whether the stock is above or below $100 a share because that was the par value. That was just like the standard value for the stock and it was supposed to be the book value, et cetera. Then people kind of viewed equity as just the most junior claimant, like the way you'd view equity, the equity slice of a CDO or something like that. And now we view equities completely

differently. And so we deploy money completely differently. Like it would be incomprehensible to an investor circa 1926 to say, I'm going to invest in this company that has no operations. It's just people. And I'm going to put the actual money into this business, but most of it will be owned by these people. And then the business, you know, my stock could be worth 50 times as much in a few years if things go perfectly. Like, that wouldn't make any sense of them. The idea of a stock going from par value to 50 times par value is just insane. So we, but that change actually was part of, I think it was like causal and in both directions with America just having a really flexible financial system. We also have a really flexible labor market relative to other country, which means we will be

patient zero for like all the job loss stuff. Like it'll hit us before it hits anyone else and it'll hit us harder than anyone else. On the other hand, one of the unique things about this particular general purpose technology rollout is that it is happening much, much faster than before. But the thing that slightly offsets that is that the specific technology actually gives you access to information and cognition such that you can ask chat GPT, like, here is my job. Like, I've been selling insurance for 20 years. How should I reskill? Like, what should I do?

And it will ask follow-up questions. It'll ask, okay, well, you know, what other problems do the companies and people you work with have? Or, you know, let's break down the skills that you have. What makes you a particularly good or bad insurance person? And then what are the other jobs that might be less AI exposed that you could do? Or, you know, as the models get smarter, you could just sell it, hey, I want you to invent a new job for me.

like a bespoke, like, this is the job I was born for. No, you already have that job. That's right. Podcasters are safe, I guess. We can ask questions. But, I mean, it does seem kind of dystopian when the upside is, well, you can ask the chat GPT, what your alternative job is. No, but this stuff always feels dystopian at the time.

Like, if you told someone 200 years ago, hey, it's going to be completely non-viable for you to inherit your father's farm and work on that farm and then give that farm to your son. And if you said, you know, not only is that economically non-viable, but you're also not going to be so fixated on is it your son or your daughter. And also, you're probably going to move to a city. You'll be surrounded by complete strangers. You'll have a job like in this loud, noisy, very uncomfortable building, like, messing around with, you know, whatever these physical things you're manufacturing are. Like, a lot of people would say that's kind of nightmarish.

And in fact, a lot of contemporary observers did talk about that being kind of nightmarish. But they did end up working out reasonably well over time. It was just new and weird and completely incomprehensible from the original standpoint. Let me ask the question in a slightly different. way, which is like, who captures the productivity gains here and how are they distributed? Because I can imagine a situation where we all have jobs. There are certain tasks in our job that we don't necessarily like doing. And if we can use AI to do them more efficiently, then maybe that's great for us. But history is full of examples of new technology that is pitched as, you know, a productivity enhancer.

Like email is going to let you do things faster. And then it turns out that actually email means we have to reply to emails 24 hours a day and we just get more volume and it actually makes us more miserable. Who captures the efficiency gains here? People who don't feel miserable when they can produce more economic output but have to pay different kinds of attention or put in more effort. It is true that these things have negative side effects, but when the cost of communication goes down, the amount of coordination you can do goes way up and you can coordinate in different kinds of in different ways. And I think this also illustrates that it's very hard to predict the

demand for service sector work because a lot of it is so meta, like a lot of this interacting with other parts of the service sector. So like Excel is kind of the tried example, but there's also word processing where you could think, okay, word processing makes it easier for lawyers to just quickly draft contracts. And so we'll need fewer lawyers. But actually it meant they could turn a two-page contract into a 50-page contract and then you need more lawyers to handle that. Joe, did I tell you, I met someone in Connecticut who was training to be a copy editor. He started two years ago. And I just thought, my God, what bad timing.

I've heard a few other stories. I heard someone recently, they went to like a coding boot camp like six months ago. I'm like, that's awkward timing. David, in your polling, who likes AI the most and who likes AI the least? Yeah, there's a pretty clear coalition. You know, the actual levels depend a lot on how you ask it. But the main demographic split is that young people like AI a lot more than older people than men, more than women, which is probably unsurprising.

And then, I think, interestingly, generally educated people have much more positive views. Ironically, despite all the talk of the white collar job loss, it's the people with degrees who I think are the most optimistic. And then working class people are a lot less optimistic. And then finally, after all of that, generally speaking, white people are more pessimistic about AI. and black and Latino voters are generally more optimistic. You know, the Mississippi Delta, for example, actually has the highest rate of folks who are excited about AI. That's interesting.

Wait, explain the difference in attitudes between the educated and more like working class. Is that just because the working class is, I guess, maybe historically more used to getting screwed over, for lack of a better word? Well, I think that is exactly it. That, you know, the way, what I'll just, he painted the story of, well,

you know, the jobs are going to change, but there's going to be tons of growth. There's going to be new jobs. And I think it's just really worth saying that voters are extremely skeptical of this claim. Like if you go and you say, oh, how much do you trust the statement that AI is going to create lots of new jobs? I think it's something like minus 40. The reality is that, you know, voters are extremely negative about the economy right now. It's really impossible to overstate where something like two-thirds of the public thinks that the economy is rigged, only 35% of the public thinks feels that they're financially secure. And in that context where people are genuinely quite angry, they are extremely skeptical of the claim that things are just going to be okay. And obviously, you know, working class people remember the decline in

manufacturing. They're like basically every economic, big economic shift has had winners and losers. And the winners generally have been either the top one or the top 10% be you've come. argue about that, but other people have lost. And so, you know, the main point I want to make is just, I think the picture that Bryn is painting isn't going to be allowed to happen because the public has a say. This is a really important point. The public will sort of, it's almost impossible to talk about the future without knowing what the politics are going to be.

Let me ask you, so you work with Democrats. And when I think about the Democrats and AI, I think there's like a few different things. things, there's a pretty big contingent on the sort of capital L left that thinks it's all a fraud, that thinks it's like, this is Theronose again, this is NFTs, this is completely economic sustainable. Then there's sort of like the anti-data center people. They don't want them in the backyard. They don't want them around, period.

Then there's sort of like what's been building in the Democratic Party for a while, just this sort of like anti-big tech and the sort of the anti- oligarchs, stuff like that. Like, is there anyone that you talk to who's most, I'm trying to think of the exact term, taking it very seriously as an important technology that is going to evolve? Like, is there anyone you talk to is like, no, this actually works, this is real, this could be productivity, gains, et cetera, or is it almost just various flavors of political negativity? Well, I think the backdrop is that this is a very new political issue. Even since last year, the share of voters who care about AI has increased more than any of the other 39 issues that we're tracking.

And so politicians obviously are catching up. Usually politicians are pretty reactive. But you know what I will say is I think that Democrats are in a much better position to capitalize on this than Republicans are, just for the basic reason that Republicans have really painted themselves in a corner on this. You know, Donald Trump is on tape saying that AI is going to create tons of new jobs, that job-lawful. loss isn't going to happen. J.D. Dance gave the speech where he was like, oh, we will never regulate AI. And so I think, you know, just for my conversations, I'm seeing Democratic politicians care a lot more about this than they did six months ago. And I think that folks are kind of ambling

about to figure out what the right way to respond is. Wait, say more about why you think AI has become a concern so quickly because, you know, everyone in this room has probably played around with chat GPT and things like that. But I would imagine for a big chunk of the population, it has hasn't necessarily impacted their day-to-day lives just yet. So it's kind of surprising to see AI concerns rise the ranks of worries so fast. Well, I do think people see the writing on the wall. You know, that basically as it stands, something like 60% of the public has used these tools. 13% of the public uses them every day.

And I think that people really underestimate the extent to which the public is concerned. or the, I think these tools really are being rolled out quite widely across a whole host of different sectors. You know, I was talking this weekend to, you know, a medical tech who was like, oh, yeah, no, the AI is being deployed. Like, you know, she was in a rural hospital in Montana. And even them, she was like, oh, yeah, no, AI is being deployed quite widely. And I think that, again, this is happening in the context of voters feeling extremely negatively about the economy. And so whenever they see the prospect of a large-scale revolution and how all of it, their jobs are going to happen. I think that they're very concerned. They're going to be screwed.

Byrne, you're plugged into the opposite side of the aisle, I believe. Yes. And on there, there's some interesting cleavages as well, because obviously we have the sort of the tech right that's very enthusiastic, the progress and acceleration and so forth. And then you have the sort of obvious, like, populist coalition. You have, like, politicians about how awful it would be if, you know, we ever had self-driving truck. and how terrible that would be for drivers. How do you see the tectonic plates moving around on the other side? Well, I like AI, so I'm very pessimistic on the political dimension.

And I think part of it is that when you ask people about AI, when you're making it salient, they have one set of views. But if you look at their behavior, they have a different set of views. And I think this is like a broader point about the large-scale deployments of technologies is that they do increase measured income and wealth inequality, but they decrease consumption inequality. So things like flying on a plane is just a much more attainable, affordable thing. There were some recent stats on how DoorDash usage is heaviest among lower income people.

And so you just have access to a lot of things where it used to be that either nobody had it or only very wealthy people had it. And actually, Chet-TPT is the kind of thing where if it didn't exist and I were much, much wealthier, I would just hire people to do that kind of thing. I would just ask them weird questions about history or just send them off on little research tasks. and I wouldn't feel at all bad if they get back to me, you know, a week later and present me with this report. And I say, oh, I actually change my mind. I don't care about that anymore. But here's the next one. But like, you can do that with an LLM. But also, older demographics nominally don't like AI, but they spend a lot of time on Facebook, which means they actually do really like AI.

They like AI recommendation engines. They are very tolerant of AI recommended ads. They love AI generated images and AI generated text. They feel much more comfortable on a site where an AI is actually going to tell them what the comment should say, and they don't have to come up with a comment. Like, people actually love AI from a consumption perspective and hate it from, like, the outside abstract perspective. So the thing that makes me more optimistic is just the salience of these different issues fluctuates

over time. Maybe AI deployment is so fast that becomes just part of the background noise, like the internal combustion engine or electricity or even the internet. We're like, the internet is not a campaign issue right now. It was sort of an issue in 2000. And it's been, I guess, you know, there were internet-y issues in. in 2016 and 2020, but it's becoming less and less salient. We just don't think about it, like, are you pro or anti?

Like, it just is. And I think that with a lot of these invisible productivity gains from AI or less salient productivity gains from AI, we're in a better world where people are not thinking AI is and is only the homework cheating and plagiarism, like denial a plagiarism machine, plus the thing that's taking away my best potential job. Canadian women are looking for more.

More into themselves, their businesses, their elected leaders, and the world around them. And that's why we're thrilled to introduce the Honest Talk podcast. I'm Jennifer Stewart. And I'm Catherine Clark. And in this podcast, we interview Canada's most inspiring women. Entrepreneurs, artists, athletes, politicians, and newsmakers, all at different stages of their journey. So if you're looking to connect, then we hope you'll join us. Listen to the Honest Talk podcast on IHeart Radio or wherever you listen to your podcasts.

Do either of you have a favorite historical analogy for AI right now in terms of understanding it from a political or social perspective? You know, for me, as I said before, I think it's COVID, where I think that this is going to hit very quickly. People get obsessed about, you know, exactly when it will happen or exactly how it will happen. But I think voters hate change. I think that that's one of the most underrated. Status quo bias of politics is very strong. if you look at who are the most popular politicians in the country, it's always the governors who do absolutely nothing, you know, underrated political fact. So I think if you're talking about a world where every single job is being simultaneously transformed and there are tons of winners or losers, like, you know, obviously people get hung up on this question of will everyone lose their job because of AI.

But if like 3% of people lose their job because of AI, it's just going to be the biggest issue in the world. You know, whenever you have diffuse benefits and concentrated losers, that's just like a public choice recipe for chaos. And so I really like the COVID analogy because COVID happened basically all at once. And then our political system was scrambled and new coalitions were created and it was very difficult to respond. And whenever you go back, as we talked about before, it's really hard to think of an economic transition that happened that quickly. Fern? Yeah, I think the COVID analogy might be revealing in another way, which is like, the coalition's completely shifted multiple times.

It was like in January, it's only the weird, online, extremely online, right-wing, anonymous people, and then a handful of rationalist people who say, this is a really big deal. And it's like, we need to shut down air travel from China. And like for the EA types, for the rationalist types, that was more like, this is a prudent response to a potential existential threat.

And then I think for a lot of people on the far right, it was like, this is a way to make China look bad. and also to have fewer flights to and from China, and therefore we want to do it. And then, you know, Trump, I think, was very just tied to what the S&P was telling him about whether this was a good thing or a bad thing. Like, you could kind of see it. I think if he had a live ticker telling him how the market was reacting to some of his early COVID speeches, we would have had a completely different COVID policy of the country. And then we kind of did this flip where now then the right became the more COVID libertarian, you know, just let it rip. we're all going to be immune to this at some point, a faction, which was really weird to me

because the right skews older and older people were more at risk. But I guess it goes back to salience and information environment. And if you are, if you're in the cohort where you're, you're sedentary, you're elderly, and you are sedentary because you're spending a lot of time watching TV. If what you're watching on TV is Fox, then maybe you just get a very different view of the world. So I think the coalitions probably fluctuate a lot. I don't think there's actually any, I don't think either major party is a good home for the pro-growth slash abundance faction. like they can be a minority among Democrats or minority among Republicans and will have to sell out in some ways to have any kind of influence whatsoever. So in that sense, again, pretty pessimistic.

But when you were talking about the issues that changed in salience, I think in the data you had, it was like a year ago, it was trade and everyone was obsessed with trade. And then trade became a huge thing and was like the main producer of headlines for a while. And then it kind of faded. So even though like the situation is still very messed up in terms of global trade. In fact, it is more messed up than it was a few weeks ago in terms of oil trade. That's right. So in that sense, trade is like the day-to-day, like, are you better off than you were a week ago question? Trade is still the most salient issue.

I mean, maybe if there's a new model, really subtle change things. But for now, that's the case. So, yeah, it's all pretty volatile. In terms of the technology that I think maps most closely to this, I think the transistor and integrated circuits are probably the closest mapping, one in the literal sense of they are a way to make everything in your life a little bit smarter. And it is just astonishing to think of how many little idiot savant devices we have in our homes that can do little calculations and computations and have a nice little interface

and how that stuff got so cheap that it basically became free. It would not make sense to think about can we build a microwave that has entirely mechanical controls with no transistors inside of it? Like, why would you do that? Speaking of politics, like, David, do you think it's plausible? Let's say the president were for it a few. future president or for it, someone puts up a bill and says, we're going to ban AI new AI data center construction. Could you see a world where that actually passes? Because I feel like

that's like that seems crazy, but also more I think about it, it almost seems like it could pass. Yeah, I could see all kinds of things passing. We're going to, we will probably have divided government and so it's always good to bet against any specific thing happening in that situation. But like we just, you know, it's very popular, for example, to ban investors from owning homes, for example, which is both parties seem to be like very into that idea that corporations shouldn't own large swathes of single family homes. Totally seems to cut across both parties. The anti-Data Center thing strikes me as similar where it does not seem like a right or left thing. It seems like a very broad, populist TikTok politics sort of thing.

Yeah. You know, what I'll say about the polling on data centers is it's true that if you poll, you know, do you want a data center in your neighborhood? People say no. But the flip side is if you add literally anything to it, if you're like, what if it's built by clean energy or what if it lowered your tax bills by 10 percent, then suddenly it becomes really popular? I've seen a lot of politicians like kind of grab for the data center thing because this is just like a big, scary issue and the data center stuff is just like a really legible over land use and this and that. But I think it would be a mistake, Not because I care that much personally about whether we ban data centers or not, but really just because I think the public really wants something else. Like if we ban data centers tomorrow, that really doesn't change the fact that voters think that the economy is rigged or that they're very scared about the future.

And, you know, in our testing, you know, we've generally tested dozens of different ways to talk about AI. And, you know, the data center stuff, I think, just doesn't move the public very... very much. While what does, I think, is actually just being a lot more radical, talking about job guarantees or income guarantees or eviction protections. Like, I think that there's just an enormous amount of fear and the political system should try to address it. And it's not just should. I think it will, you know, because we do live in a democracy. Actually, Byrne, that reminds me. One of the things we hear a lot when it comes to AI is this idea of electricity is a limiting factor. right? Everyone talks about how that's the big constraint. When you yourself think about adoption of the

technology and it's spread across America and I guess around the world, how much are you actually thinking about things like data centers and electricity consumption? Well, I'm thinking less about those right now just because everyone who is in that supply chain has all their capacity booked out so far in the future. Yeah. Like new supply is pretty inelastic. Like I wouldn't be entirely surprised to see the AI companies just guaranteeing that they will take delivery on, you know, something from Siemens Energy or G in, you know, the year 2032 or whatever, just so that they break around on new manufacturing capacity. But I kind of view that because it's such a long lag thing.

It's kind of a given. I think that the actual constraints are more on the internal, like, organization side. Just, you know, do you even have an org chart if you have AI everywhere that's always routing messages to whoever needs them. Like, could you just have, there's one person who's the CEO and everyone else is an individual contributor and you AIify all metal management? Probably not, because the other limiting factor is just liability is that you do want a human in the loop. I was joking about this earlier, but it is true. The person who's in an ideal position right now is someone who is in a regulated job where there is some kind of trade association that limits new entrance into that job,

but there's excess demand for the services they provide, then AI can increase their output per hour. These people will be minting money, and they will continue to do so as long as they can limit supply inflows. And so what we might end up with is a more kind of guild-fied economy where we have limits on who specifically, not who can do the job, but who can actually stamp it and say, I'm taking credit for this and you can sue me if it goes wrong. Because that's actually a really valuable thing. And it's still, like, you can't really sue a data center. You know, you can try to sue Open AI, but they can probably afford better lawyers than you. You actually, it's probably more economically efficient to sue me for something chat. GPT told me to do that I didn't double check by asking Claude. And so we'll, like, it is weird to think that

one of our economic purposes as, you know, living and sold human beings is to be an easy target for a lawsuit. But it is actually something economically valuable that we can do. And because we all have different risk preferences and risk tolerances for things, it actually means that we would get this sort of uneven deployment of AI where there would be people who just demonstrate that you can go way too far and that if you decide that you're going to be the CPA who does taxes for, you know, does a thousand, 10, 40s a day, you're probably the first CPA to go to prison specifically for something that an LLM told you to do if that hasn't happened yet. But I think that that kind of model where you want a human in the loop because so many structures that we have economically and socially

just assume there's a person that will probably stick around. You write a lot about finance as well as technology. And I've been thinking a lot about the future of finance and AI. One of the things that comes up is we did a really good episode with the CEO of PNC Bank. We were talking a little bit about AI and lending. And he said, you know, if you deny a loan to someone, you have to have a reason for it. And there are various laws that, you know, anti-discrimination laws and stuff. So if someone has denied credit, you have to be able to explain why. One of the things with AI is that it can make very good decisions, but it's hard to interpret. And the AI often can't explain how it arrived at a conclusion. And I'm curious, like, from a, just thinking about the finance

industry, how much do you think that's going to be a limiting constraint on the degree to which AI disrupts the industry, the fact that a lot of things need to be articulable in English, basically? Well, I think AI, like humans, is bad at knowing why it actually did things and really, really good at explaining why whatever it did was the right thing to do. So it probably, I mean, I'm not the head of PNC, so I don't know for sure, but I would imagine it actually makes it easier to just come up with some very solid-sounding rationalization for anything. They're just really good at rationalizing. So I would be less concerned with that. Like I actually think it's nice to have more open, accessible kind of reasoning.

Like you can actually read the reasoning traces. Yeah. And one of the effects I think it has on finance is that people who are coming up with recommendations like make this loan, don't make that loan, it will make a lot of sense to have much more detailed records of their entire thought process. So you might have them working in Databricks notebooks or something equivalent to that, specifically so you could see, okay, first they asked this question, and then they went down this rabbit hole, then they decided it's irrelevant, and then they went to this question, and then they spent some time on it, et cetera.

Because if you have just, here's the question and here's the polished answer, someone can happen with and edited, you're missing a lot of the intermediate layers. And so it's hard to train a model that can trace through that and reproduce it. Now, if you have a smart enough model, it's basically implicitly reproducing all of that reasoning on its own. But that means that it does worse when the reasoning is really, really clever, and the person didn't show their work. And a lot of clever people just figure things out, and then they're already bored with the problem. So they don't want to tell you how they did it. And they move on to the next thing.

I did have like a home DIY project recently. And I was asking chat GPT how to do it. And it basically told me to defy gravity and could not explain why it had to how it arrived at that conclusion. I'll tell you more about it later because it's a long. story. But anyway, if we could do a little bit of like political polling, naval gazing for a second, David, when I think about AI and some of what you do, I think it could be very helpful to you because you can identify even more granular issues for the population. You can come up with like even better polling questions. But then I think that everyone's going to have access to basically

the same technology. And so the worst case outcome is probably going to come to fruition, which is we're just going to get more identity and sort of cultural grievance politics. How do you see that going? Like, does political campaigning get smarter with AI, or do we just kind of descend more into culture politics? Well, obviously, it's very hard to make predictions about the future, which I guess I've said already today. But it's easy to imagine really bad things happening, deep fakes, the inability to track what's true and what's not, you know, the ability now of lots of people to make content that's persuasive that argues for things that previously weren't within the domain. But, you know, I do think in this discussion, people kind of underestimate how dysfunctional the status quo is, where if you look

at social media, for example, something like 5% of the public is responsible for a majority of social media content, which is crazy, you know, and I think that right now, because content is expensive to produce. You know, if you are a influencer or a writer about politics, your economic incentives are really to focus on the 5% of the public that is consuming way, way more political content than everyone else. And really, basically, no one in the political spectrum right now is making content focused on regular people who don't care that much about politics. You know, just to give an example, someone I know had a panel of TikTok users where he, you know, recorded their phones and it was like 200 people and he was paying them to do that and he could see

who was watching what and after Charlie Kirk got shot there was one person who was responsible for a majority of the Charlie Kirk videos just the at day he was like really swiping and watching hundreds of Charlie Kirk assassination videos and you know if you are a content producer that is currently your incentive and so my point is really the status quo is really quite bad and I think that if you look at who these people are they tend to be quite anxious they can't They tend to be quite neurotic. Like right now, the attention game really pushes you toward being more negative, while the persuasion game of like, how do you actually get someone to change their mind really pushes

you in the opposite direction. You know, every time I've done a poll, every time I've done a test, it's generally said you should be more positive. You should focus more on regular things that affect people's lives. And, you know, the reason why that doesn't happen are the incentives of the actual content creators. And so I could see, you know, lowering the costs of producing content and kind of broadening out who is able to make content. It could be good, but also, you know, to be clear, there's a lot of ways it could be bad. We're just entering, I think, a totally different world where

it's hard to predict anything. It is weird how much like politics, discourse and ideas truly seem disconnected from anything that affects people, many people on a day-to-day life. I mean, this has been brought up before. Why has no politician really made a big issue about like getting rid of like spam texts or something? Like we all find it incredibly annoying. And yet you wouldn't, it's almost unimaginable. I'm sure you could find some one issue voters on spam texts. Right. But like that would be great. If someone actually like, let's take this seriously as a thing that annoys everybody and let's try to make a push. And yet it doesn't happen, does it? Yeah. The stat I really like on this is if you ask people, what issue do you care about the most? It's cost of living by an enormous margin.

But the flip side is that if you look at the 0.7% of the population that is donated to a Democratic campaign in the last year, then cost of living goes from number one to like number five and like climate change is on top. And so I think, you know, if you look at, like I think the status quo is that politicians right now, if you look at what they talk about, it's much more explained by what their donors care about than, you know, what the public cares about. And that's really pretty bad. I think an enormous source of our political dysfunction right now is really that people on both sides aren't listening enough to what regular people care about. Do either of you have a good read on the policies you would expect to be actually politically viable

if we do start to get, I'm not going to say an AI doom case scenario because I know you disagree on that, but a painful AI adjustment process. I think that the public is much more radical on this issue than people think. Me personally, I'm not usually the person who goes out and says, ah, the people crave radical policy change. But we really are in a very radical time. Right now, if you ask voters, you know, do you support price controls? The answer is yes by two to one, which is not something that was true five years ago. You know, when we've done tests, you know, we had one test where we had a really quite radical thing where we're like, oh, we're going to guarantee your income up to $150,000. We're going to make sure that you have a job. We're going to make sure that you won't

be evicted. And it tested better than like 98% of the clips that were made, but that were made by Democratic professionals. That specific policy, which I think is quite radical, I think is something like plus 30 and plus 15 among Trump voters. And so that's just the thing I want to say is, you know, right now everyone has this discussion about timelines and policy specifics. And, you know, the public isn't a much more radical place than politicians or commentators are. So that's, I think, the big thing I'd say is I think it's the most underpriced issue. The very best testing topic, better than populism, better than AI, is AI populism. And I think that's the direction things are going to go. Burn. My thoughts is, this is ominous. I did want to return to two of the points that

you had made on deepfakes and on how everyone in media is writing for the tiny minority of people who really care about, or everyone in political media is writing for this tiny minority that cares about politics. And one, I actually, I do think deep fakes are a net in the specific sense that... This is great. We get to a whole hour on this question. I've been writing about this one for a long time. My view is that it is always possible to create manipulative cuts of some media.

It's always possible to say, okay, there have been, you know, 50 videos recorded in the last day of some egregious happening. Here's the one we're choosing to make a new story. That's always been possible with sufficient resources. And now it's possible for everyone. And so to the extent that when you choose what narrative, if you're in a new story, if you're in a position to choose what narrative is promoted. I think of it is kind of the modern, it's like it is the de facto electoral college, is that people opt into viewing certain

kinds of media that will then shape their views to make them more correlated with that kind of media. And so what we're kind of doing is democratizing democracy. We're saying anyone can make a misleading video. Like you don't have to watch 20 hours of footage of this person talking to find the one gaffe. You can just fake it. And the fake is actually, like for the fake to be a plausible, you know, opinion shifting fake. It has to be something fairly similar to reality. It can't just be like Donald Trump is actually an alien. It has to be something like Donald Trump took a bribe in cash from a Qatari businessman or something like that.

But then if you can only make fakes that are actually kind of plausible, then they just exist in this space that is not that far from reality. And similarly, like the big viral videos, they are often a sample from reality, like they are a sample from reality, but they are something where you can magnify the perceived. frequency of an event if there's just wall-to-wall coverage of that event on video. So that was one point, I do think that it actually just makes us a less, a culture that is less likely to make up our minds on important issues by watching a, you know, grainy, shaky 12 seconds of footage of something. I think that's good. I think we should read more and watch less.

But I also think that when the, you know, the other piece of AI, just the targeting recommendation algorithms, there's a huge chunk of AI spending and AI's impact, that probably does mean that it's potentially more likely that someone could make a career out of something like agitating against spam tax. And we did actually see, you know, Click to Cancel is kind of moving along. Yeah, that's true. There's little quality of life things. Like click to cancel is the kind of thing a mayor would be really proud of, but it has to be done at more of a federal level. But maybe if it's possible to actually target whatever population niche, like really wants to bring back pay toilets because they think that banning pay toilets creates all kinds of perverse

economic outcomes, which it does. You can find those people. You can mobilize them. And you can get them to, you know, if there is like some very close race somewhere, in the House race, for example, and one of the politicians happens to support this particular pet issue, they could be like the crypto people and just money bomb the person who happens to also support the thing that they like. So we could actually have, that's another way we could potentially see AI democratizing democracy further, is that you can actually coordinate interest groups for people who are just less politics-brained, but still actually care about problems that should be solved through the political process. I just wanted to jump in quickly.

Sure. You know, you said that, Bryn said that this is ominous. And, you know, I want to push back on that a little bit. There's an Ed Glazer quote that I really like, which is, you know, everyone wants macroeconomics to have micro foundations. But micro foundations itself should be microfounded. That we are not going to experience any of the potential productivity gains or we'll only experience a fraction of the productivity gains unless we can give voters. some sense of security, that public is really crying out that they want economic security and that they want to be able to look out at the next five years without fear. And if you can provide a new

social contract, if you can actually provide economic security at a high level, then you can actually have all of these like sector-specific shifts to have higher productivity. But if politicians don't advance a vision like that, then we're just going to collapse into Byzantine, series of sector-specific, you know, regulations and guilds, which I think you don't want either. And so I think, you know, the libertarians should really choose, you know, what world they want. You know, either we can have a large-scale solution that protects people's incomes and prevents there from being losers, or we can just kind of have this giant negative sum fight playing out in every single sector simultaneously. And I'll also say, you know, we're not really in the ideal

political circumstances, you know, for that latter thing to happen. And so I think it really could get quite ugly. David and Byrne, thank you so much. That was a fascinating chat. We really are going to have to do another hour on why deep fakes are good. Thank you all for joining and have a great rest of your day. That was our conversation with David Shore and Bern Hobart, recorded live at South by Southwest in Austin. I'm Tracy Alloway. You can follow me at Tracy Alloway. And I'm Joe Wisenthal. You can follow me at the stalwart. Follow our guest. Byrne Hobart, he's at Burn Hobart, and David Shore at David Shore. Follow our producers, Carmen Rodriguez, at Carmen Armin, Dashoban, Dashobatt at Dashbot, and Kail Brooks at Kail Brooks.

And for more OddLots content, go to Bloomberg.com slash Oddlots. We have a daily newsletter and all of our episodes. And you can chat about all of these topics 24-7 in our Discord. Discord. G.G. slash OddLots. And if you enjoy OddLots, if you like it when we record these live episodes, then please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, you can listen to all of our episodes. Absolutely ad-free. All you need to do is find the Bloomberg channel on Apple Podcasts and follow the instructions there. Thanks for listening.