Loading summary
Transcript114 lines
- [00:00]
A
Foreign.
- [00:08]
B
Welcome back to Firewall. I'm your host, Bradley Tusk. With us today is our friend, my partner at Tusk Ventures, and frequent guest Bob Greenlee. Bob, thanks for coming on.
- [00:20]
A
Thank you for having me. Great to be here.
- [00:22]
B
So the topic of this episode is AI regulation. And the genesis was I'd written a substack a couple weeks ago about sort of a taxonomy of how to think about AI regulation. And then Bob sent me his taxonomy, and they were different enough that between sort of the conceptual question of how do you even go about something like this, and then the practical, you know, what do you need to do and who's doing it well, and who is it and whatever else, it was enough, I thought, to make for an interesting episode. So that's what we're going to talk about. It's okay. I think I'll start off by just laying out the taxonomy as I see it. And then, Bob, you do yours, and then we'll start debating it. So the underlying point here to me is that I have been in Politics now since 1992, and I have never seen an issue like this in that it is so pervasive, not meaning, like, media attention, but just meaning in all the different areas of society that it can and will touch. And therefore, when you think about regulating something, typically it's a specific issue. It is health care, it's education, it's energy, it's whatever it might be. And you're working within the construct of that regulatory framework. And then if you think about our work on the venture side, it's typically a portfolio company that is doing something that hasn't been done before. And then the question is, how should it be regulated? And you have the incumbents and the insurance interests usually trying to stop it. We're trying to move it forward. Sometimes there is no regulatory framework for it at all because it's a white space area and you're trying to create the right one. So but even then, it's still underneath the overall aegis of a specific vertical and topic. And the thing with AI is it touches everything. It touches consumers, it touches businesses, it touches homeowners. It impacts our energy system, it impacts our health care, our education, our economy, our safety. It involves catastrophic risk. It has a huge impact on unemployment numbers and the job market. It has the ability to do a lot of good. And so the way that I structured the taxonomy was four categories. The first is consumer protection. And this one is sort of the easiest of the four, because we already have a system of consumer protection that exists Mostly at the municipal and state level, sometimes at the federal level, depending on the agency, where the idea is basically as you would imagine, a company or an entity is doing something, and if it is seen as harmful or fraudulent in some way, the government stops it or creates rules around it. So I see a couple of areas right now where AI is being regulated from consumer protection standpoint. First are chatbots. I think we've seen about a dozen states this year pass or about to pass legislation. And it's all kinds of things. What can and can't chatbots do. Preventing sort of sexually explicit content disclosures they have to make, providing mental health advice, especially to teenagers and things like that. And those have been regulated and on a very bipartisan basis, both in terms of the votes in the states that have done it, but also the types of states, meaning both red and blue. Second is a topic we've talked about a lot on this podcast. Bob and I have talked about this on this podcast before, which is energy costs for consumers thanks to data center energy consumption. And it seemed almost inexplicably that the plan of the hyperscalers was just to build these data centers and plug into the grid and, you know, let everyone share the cost. Which is insane, because why would your average homeowner want to have a 30 or 40% higher electricity bill, especially when they're already worried about losing their job because of AI, simply to help Sam Altman become a trillionaire. Right. Like that doesn't make any sense. And so again, you've seen bipartisan support within states, and then both red and blue states, and I think we're at something like 37 states now have at least introduced legislation to regulate energy costs, typically meaning that they can't be passed on to everyone else in the grid. That could be through providing on site power. It could be through special types of purchasing agreements. It could be through using more energy efficient forms of compute, which is something that Bob and I are working on a lot. We can get into that. Third in consumer protection is around hiring and the use of AI in making hiring decisions. This has been on for a while. So at Touch Strategies, we passed a bill for a company called Pymetrics, which was an AI hiring company. They've since sold. I don't know if they're still using the name Pymetrics or something else now. And it was in the New York City Council. The bill was around kind of use of racial demographics. And the reason why we wanted it was because our biggest competitor didn't have that particular functionality. It was a very controversial bill, but it became kind of the basis for other states. So that's consumer protection. The second for me is catastrophic risk and harm. And it's tricky because, you know, states are doing it because the federal government has yet to regulate Internet 2.0 social media, let alone AI. So they're taking the lead, and I certainly don't mind that at all. So California, New York, Colorado have all passed bills that deal with frontier models, and I think they're decent bills. You know, we worked on the raise act here in New York, but, you know, the catastrophic risk of AI seems to change constantly. You know, Claude sort of talked about Mythos a couple of weeks ago. That felt like a totally whole new level and change in the potential risk of AI. Mythos is supposedly able to sort of hack anything. And so, you know, it's one of those areas where it's not just a question of whether you have regulation or not or a partisan view on it, but you need sort of a living, breathing structure, I would argue, that can evolve as the technology evolves. And, you know, the catastrophic risk is very high, whether it is AI, you know, controlling missiles that it could launch without any sort of human authorization. AI either, you know, figuring how to construct a bioweapon or enable, you know, crazy person to do so or anything else. Third category for me is jobs. In this case, it's more about harm prevention. Again, kind of like catastrophic risk, which is what we're already seeing significant layoffs because of companies using AI. Facebook just announced or Meta, that they're going to lay off 10% of their workforce because of AI tools. And you can't really prevent. I know the left will try, but you can't really prevent companies from becoming more efficient. Right. That doesn't make any sense. So the question then becomes, what do you do with all the people who lose their jobs and how do you help them? The typical response that I hear from politicians is job training, which is their answer for everything. But the problem is training them in what? You know, I'm an early stage vc. I see all the companies that purport to be the company's industries of tomorrow. And I couldn't tell you what you would train people in. And we're not going to have 200 million plumbers and H vac technicians. And so the only good idea that I've seen, and I want to talk to get Bob's take on this is our friend Daniel Schreiber, who's going to come on the podcast at some point, talk about this is the co founder and CEO of Lemonade, an insurance tech startup that we invested in and worked with. And he was in the, he lives in, in Jerusalem, but he was in New York A few weeks ago we had lunch and told me about a white paper that he had funded with the idea of taxing companies on their incremental profits from fewer headcount and then redistributing those profits through. He called it the negative income tax. You could also call it universal basic income to the people who were laid off. And his argument was it kind of works really elegantly because as the layoffs go up, so does the revenue tax revenue and therefore the ability to take care of the people who were laid off. So that's three and then four for me was all the good things of AI, right? How can you use AI to make government work better and make society work better? And I gave two specific examples of our portfolio. One is going to be called Doctronic. Thanks to Bob and Marlo Kanemitsu, they received the first license ever to do prescription of medication via AI. This is in Utah. And if you think about it, you have presumed for a doctor, you need a refill. It's a pain in the ass. Either you got to track them down and call them and wait for them to call you back or set or hold for 30 minutes or you have to go to an urgent care center to kind of get something, just have them do it instead. And it's just a huge waste of time and productivity. And in a system where affordability and accessibility is by far the number one issue, you know, this improves both of those. So there are things like that. Another one's called Hazel. So they do procurement right now for government procurement they're working on, they use AI for that. They're working on private sector verticals too. And you know, the way it works is imagine you're a school district, you need pencils. Now you run low, someone has to notice. Then they have to tell someone else, who tells the legal department, start drafting rfp. That takes however long it takes. They review it, someone has to sign off on it. You're waiting around, it finally gets sent out. But it goes to like the eight pencil vendors they know those pencil vendors could easily be doing bid rigging or anything else. Then when they're responded to, however long that takes goes back to a committee of people who if they're corrupt, mean they're just stealing from the taxpayers outright. But if they're honest, which is the case in my experience, most of the time, they're still human. So they have bias and friendships and rivalries and all this other stuff that gets in the way. And it's just a wildly inefficient system that ends up leading to way higher cost for taxpayers, a lot of wasted times. I mean, more bureaucrats than you need on the payroll, and sometimes corruption itself. And so areas where I can help people. So to me, those were the four categories in the way to think about this. Bob, you approached it a little differently. What's, what's your take?
- [10:26]
A
So I, I think I looked at it from a two part lens. The first part lens is like, how do you think about the taxonomy of the government process generally? And then second, how do you think about the taxonomy of the political process? So it's not, I don't think that the things that you're talking about, the areas that you're talking about are not the right areas. It's just the way I would look at it is, number one, this is just the governmental problem. Is there a problem? If there is a problem, yes. AI has problems or opportunities, right. Is there an opportunity? Number two, can government deal with it? Right. So, okay, let's talk about consumer protection. Can government help consumer protection? To some extent, yes. Right. To some extent that helping with it creates its own externalities because you take away anything you do to protect consumers, takes away choice and independence, and you reach a point of, I guess what I would call ideological equilibrium. Number three is government dealing with it. So consumer protection right now, every state and the federal government has relatively fulsome consumer protection laws. The question you would have to ask are, are the problems that we're talking about not covered by existing consumer protection laws? Arguably they are covered. I mean, and the laws that we're seeing getting passed are arguably laws that are getting passed for the purposes of making lawmakers look good rather than doing anything new. On the consumer protection side, there's a lot of laws that get passed right now that say we need to disclose to people that what they're talking to is an AI that's good, does no real harm to anyone. Arguably there's no social benefit because 80 to 90% of the people using AI in fact know they're using AI. They, you know, they knowingly went to a website and they're knowingly dealing with it. But to the extent that in the future this is a problem. Sure. Okay, if you get past the, in my rubric, if we get past that, are we already dealing with this? Then the next question is, what is the solution and depending on what the solution is. Last question is, can it get solved by legislation? Sometimes the issue is an executional issue and even the legislation won't solve it. So I'll give you an example from one of the ones that you're talking about, job training. It's not clear. A, it's not clear that we can throw money at job creating to solve it, but B, it's not clear that the policy is really the problem. Obviously the states and the federal government have had a policy in favor of training people who are disenfranchised by technological advances or disemployed by technological advances for decades. The problem is executionally, people really don't know how to do it, as you said. So it's not clear that solving new, that this is a policy problem that you can change with the law rather than an executional problem that you can change with new technologies or new ways of doing things. And then the question, I guess when you get to the how do you pass? It hits my other second taxonomy, which is the political process of what can actually get passed. And on that I look at it and say there's really three things they can get past. Things that are easy that there's really no opposition to. So on the AI side that's like election deep fakes because no elected official wants themselves to be the subject of AI fraud. So that's. That has been a very easy pass. Child sexual exploitation. No one is. No one is in favor of child sexual exploitation. Easy pass. Second thing that can get passed, things that are legitimate crises. That is, that's where I would put your externalities over electricity bills right now. When people see their electricity costs going up by 10, 20, 30%, they are in crisis. This is something that they cannot live with. People living paycheck to paycheck can't afford it. And a solution has to happen now at that point. These are the bills that, generally speaking, on the tusk side that we often take are things where issues really have reached a crisis point and something has to be resolved. If it's not technically a crisis, those are the ones that we have trouble winning. Because if people don't feel like it's a crisis and it doesn't have to be resolved, then it won't. Third area is those places that have reached what we call the equilibrium where the opposing forces on something will negotiate over time. A policy consensus that takes a long time. Sometimes you never get there. This is why section 230 has never been amended, despite the fact that everybody realizes that there are problems with it. And you will see over the course of time that that's just a repeated game where the same policymakers, let's say in the AI side, the trial lawyers versus the doctors, fight over issues every year and really just negotiate where the trenches are along the legislative battlefield. I think that's my taxonomy. A lot of the stuff that you've talked about fits into there. There are some of the things that are complicated, and one of the areas, I guess I would say is particularly complicated on the AI side is. And do if you realize that the job displacement also has revenue impacts on the state, where you can't really throw money at things if you're losing all your tax revenue because your people aren't employed. So I do think that there are, there are real externality issues you have to talk through on some of this stuff. But that's, that's how I look at it.
- [15:43]
B
Okay, so then let me sort of take it to the next level, which is take the catastrophic karma question, which I know is the most complicated one, and then run it through the taxonomy you just gave to analyze it and tell me how it should be dealt with. Sure.
- [16:00]
A
So let's look at it. Number one, is there a problem? Yes, everybody could die. That is unquestionably a problem. Is it an easy solution? Absolutely not. There are. Really, there's. Number one, it is not. Let me. I'll go through my full taxonomy. Can government deal with the issue? Yes, government could deal with the issue. They could deal with it in a very blanket way through just saying, look, we can't do anything along these lines. You can deal with it in a nuanced way by trying to do the types of things like the Raise act, where you're raising real frontier issues. The challenge is always like, the nuanced ways may not actually solve the problem, but honestly, the blunt force or the brute force issues may not solve the problem either, largely because every state and even the federal government has limited jurisdiction. And anything that we're doing can be undone by China, by Iran, by North Korea in a heartbeat. So to some extent, every choice we're making has to be ultimately something that you could enforce unilaterally everywhere and then
- [17:03]
B
also add in, and this may fit into a different part of the tax economy. But the valid concern and also inevitable argument from the hyperscalers that because of that regulating them just puts them and therefore the US Economy at a competitive
- [17:18]
A
disadvantage, or that they'll move overseas. And this, this hits that revenue question of, okay, well, let's just let you know, to Daniel's point, let's do a negative income tax on the profits made. Well, that's fine until everyone moves then to Ireland or to Dubai. Is that a real risk in the US or to another beneficial jurisdiction?
- [17:37]
B
I could see it in Israel. But is that a real risk in the US that people.
- [17:41]
A
That a technology company will move its data centers to a place where it argues it doesn't have tax risk? Sure, it's a complete risk. The answer to it, and this is the. I think the challenge, this would be my challenge to Daniel when he's on is it is very hard for the mom and pop store using AI to that touches real people in a real state, like the application layer. It's very hard for them to escape jurisdiction because they're touching real people that are in a jurisdiction. It's very easy for the frontier models, the OpenAI's and Anthropics, to move to a tax favored jurisdiction. They can move their servers, they can move their people, they can do whatever they need to do to move their intellectual property. So the question is, if all you're tacking is the application layer and you can't tax the development, the frontier model layer, are you going to be able to successfully actually capture the wealth or is all the wealth going to get siphoned into that place that's unreachable from a tax perspective?
- [18:41]
B
Okay. All right.
- [18:42]
A
I think it's a really hard question, honestly.
- [18:44]
B
Yeah, so go. But go back. You were still on the first part of the.
- [18:47]
A
Yeah, no. So can government solve it? Critical risk? I think I would say yes. Are we solving it? The answer is, I think to some extent, yes. There are things we're doing to solve, but I think there's certainly at the catastrophic risk of AI. There is certainly more to be done to solve it. And I think this is a place where potentially new legislation would be warranted.
- [19:09]
B
Federal level or state?
- [19:12]
A
Probably federal. I mean, because a particular state has limited jurisdiction, it can only solve the issues in that state. And the catastrophic issues are issues like by definition that will not be limited to a single state. Right. If, if a, you know, a, if Mythos creates a bug that hits everywhere, right. That exploit is going to literally hit everywhere. If it creates a virus that, that creates another Covid that's really going to hit everywhere states. A single state cannot solve that issue.
- [19:41]
B
So two around this one, which is one, I have a most likely naive and hopeful thought, which is that so Democrats win the House, maybe the Senate, but I'd say probably not, but it sure doesn't really matter for purposes of this question. And everyone in Politics realizes that 2028 is going to be the AI election. And AI is already extremely unpopular. If you look at most polling. And because everyone cares more about self preservation than anything else, both parties and even the President says we gotta at least tell the taxpayers we did something. And therefore in 2027 or 2028, but ideally 27, they actually do come together and work on a framework that makes sense, which I think might be politically doable, because if you do look at the issues that are happening and moving in the states, they are very bipartisan. Right. It's less ideological. Now, you do have the risk of the fact that the Trump White House can easily be bought off by, you know, anyone, and that those same companies can have these giant super PACs and try to intimidate people. Although we're seeing here in New York, in the one of the congressional districts, that the attacks on Alex Boris, who was a sponsor of the Raise act, are actually backfiring and helping him in the primary. You think I'm crazy or do you think there's some shot that there actually could be a bipartisan piece of legislation, like a landmark legislation?
- [21:09]
A
I think you're crazy.
- [21:11]
B
Okay. Why?
- [21:12]
A
I think, I think politically you're right, that there's enough of a crisis and there's enough that people will find that will have an urge to pass something, but I think that something that they pass will not be landmark. I think there's, I think there's not enough of a consensus. There hasn't been time or the tech and the technology has been moving too fast to reach consensus. So I think what they will do if they're able to do anything, and God knows they've been trying to pass crypto stuff, which is super easy in the grand scheme.
- [21:40]
B
That seems like it's finally going to move.
- [21:41]
A
But, yeah, the Clarity act, you know, but that's, that's five years of work for like, you know, effectively digitizing dollars, which is something that everybody does all the time with their cards.
- [21:52]
B
Right.
- [21:53]
A
That's taken five years. I don't see landmark legislation, which was your question, being passed in a year. I see easy stuff that they can go out and sell being passed.
- [22:03]
B
Right, right. Kind of like how prediction markets, they'll say they've dealt with it. And by the way, I'm not sure what other legislation would be needed, but, you know, they're gonna have some, like, ban of insider trading. That's, like, good and obvious, but not really meaningful.
- [22:18]
A
Yeah, that doesn't solve the fundamental issues around prediction markets. Meaning that a. I mean. Well, a. It depends what you think the problems with prediction markets are.
- [22:28]
B
Again, I'm not even sure that you need legislation beyond that, but it just, It's a. It'll be like a kind of an empty calories type thing where it looks like they did something, but it doesn't really.
- [22:35]
A
I think there are. I think there are real things that you. The prediction markets do have as challenges that we would need to think through. I mean, kids getting on youth access that you could solve with some type of biometrics.
- [22:46]
B
Yeah, that. That could be an. That's not a hard in.
- [22:49]
A
Like, you know, the revenue issue for states around gambling is, you know, if you think the solution is states should take their medicine and shouldn't have been relying on gambling revenue in the first place, which is a legitimate position, but probably not one I share, then okay, fine. If you think states needed the gambling revenue and that we're now gonna have education problems because they have huge budget holes, then you have to figure out how to solve that.
- [23:13]
B
Well, and not just that. I know we're taking this a little bit different direction, but as you and I have discussed before, if the supreme court upholds the CFTC's jurisdiction of prediction markets, which I think is at least 50, 50, if not slightly more likely than not, you would imagine that most of the sports betting companies that are licensed by states and pay taxes into states shift into just being prediction market companies instead, creates a huge revenue hole for the states, which gets filled by more types of legalized gambling, actually, not less.
- [23:43]
A
Right.
- [23:43]
B
Yeah. So I think that. And then on a corollary there. So obviously we worked on getting Governor Hochul to require in New York that biometric screening be used to log on to sports betting apps. You know, you and I would both like to see that replicated nationally. You know, it was a lot easier for me to do in New York if I had the ear of everyone already. So do you see a world where, and maybe past the Supreme Court decision on Paps makes this impossible, but. So let's say there's a prediction market bipartisan bill that bans insider trading, and let's say it also includes biometric screening to log on, which I just think it sort of seems like an easy give for the prediction markets to do. Could they create still some sort of incentive, positive or negative, for the companies that are operating at the state level to require it as well, like an additional tax federally? If you don't have this or something you can give them they want. I'm not sure what that would be.
- [24:46]
A
But do you think that's feasible for biometrics? I think it's feasible, but for companies
- [24:51]
B
that are completely regulated by states and the state authority is unquestionably the case based upon the Supreme Court decision?
- [24:57]
A
No, I think it's reasonable. I also think this is one of these issues where technology is kind of starting to make the problem disappear, the political problem. Because I think at the end of the day you're starting to see wagering companies have enough problem with AI deepfakes and AI, you know, AI exploits that they need biometric verification so that they don't. So that they're not creating, they're not dealing with a million fake accounts.
- [25:25]
B
Right. They're not going to scam themselves.
- [25:27]
A
Right. So at some point they actually need the technology too and they need to be part of the solution in terms of, okay, let's see something that works for legitimate businesses and actually control some of this can. So it can't. Can Congress over time regularly if they
- [25:42]
B
said this is a safety consumer protection issue and let's say the companies didn't want to do it, but let's say Congress said that given PABSPA and the 10th Amendment could. Does Congress still have the jurisdiction to require something like that? They do. Right. I mean states the rules of driving, but NHTSA sets the fundamental structure safety standards for a car.
- [26:01]
A
Yeah. I would look more at like FTC and states can make consumer protection laws, but you have federal advertising laws that you know, deal with false advertising. Yeah, all the time. So ftc, FCC have significant jurisdiction where you could do things like this from a consumer protection framework. Obviously there's maybe some commerce cause issues, but these don't hit the commerce clause issues would come up particularly if a state or states felt like they were intentionally singled out on an issue. But I don't, I don't see that in this context.
- [26:34]
B
Right. One more on this topic before we go to the next part of the taxonomy, which would be given that the tech is changing so fast, let's assume that everyone was really well intentioned and just everyone collectively wanted to do the right thing to protect people. The normal structures that we create to regulate things, I don't know that they work in this situation simply because it's usually something's created, then the regulations occur, the regulators get tasked with doing it and then there's an assumption that's going to remain relatively static. And here because it's moving so Fast. I almost feel like it needs a different underlying regulatory approach or structure to effectively do the work itself.
- [27:22]
A
Is there a question there?
- [27:23]
B
I guess the question is how would you do you agree? And if you do agree, how would you do it?
- [27:28]
A
So I do agree with you that this goes back to the taxonomy of the second question. Can government actually deal with it? And not from a policy standpoint, but executionally. I think one of the biggest challenges with AI generally is that government has historically had a human capital problem. You cannot hire the best and the brightest in government for a lot of reasons, like locationally, you know, say what you want about Albany, but New York is nicer. Right. You can't pay the same salaries except for the head football coach of the largest, you know, school.
- [28:01]
B
They make more than anybody.
- [28:03]
A
They are, they are always the highest paid employee in the state, except for in states like New York that don't have like a single large football team. Right. But it's almost always a head football coach. So you have these, you have these huge human capital problems that you can't solve at scale. And AI is making those human capital problems honestly worse and not better. There are obviously procurement challenges that would.
- [28:25]
B
Well, why is it worse? I mean, couldn't, couldn't I make it better? Because you could take mediocre talent and supplement it with AI to do a better job or replace it.
- [28:37]
A
I think this is, I think this is one of the things about AI that is most often misunderstood. Right. Which is AI works better with great talent than it works with mediocre talent. Feeding it prompts and doing the inferences and the like. This is one of the reasons why I think this idea that like, truly exceptional people will never be hurt by AI is very true. I mean, you will. Your people who are really great, who are great critical thinkers, will get more out of AI than people who are average.
- [29:06]
B
Yeah. So two. It's funny, two points on that. Which is one, they'll take Hazel, which is our company. Right. Procurement's a pretty well established, well understood process. They've built a system that kind of works within that. Do you need exceptional talent to use Hazel to make procurement a lot better?
- [29:26]
A
I think procurement can be made a lot better. I think permitting can be made a tremendous amount better and much quicker and much less costly for people. So I think there are tons of places in government where you can see it being better. Where I think it's really hard for government to get better is regulating AI, because I think that, I think, you know, in procurement, at like, procurement, okay, you make it better. The response from the other side is to use AI to respond to procurement. So AI is talking to AI. Same with permitting. Fine. That's a net win for everybody. But if you have the best and brightest on the okay, how are we going to get around regulation or optimize regulation to our benefit? And then you're stuck with the people actually regulating being kind of in the middle. Not bad smart people, but not like your MIT PhDs in neuromorphic engineering. Then you're going to have. You're going to be behind all the time. And it's not at all clear to me that you actually can regulate it effectively.
- [30:24]
B
Quick tangent. Which is your point of critical thinking. So I've long held the view that if you are not going to practice law, you shouldn't go to law school. It is just too many years too expensive unless you have a rich parent and a lot of opportunity cost. And therefore, obviously, if you want to be a lawyer, of course go to law school. But I get the question all the time, you probably do too, of people asking you if they should go to law school where they want you to say yes, even though when you ask them, well, do you want to practice law? They say no and they want you to say yes because it'll make their parents happy or it'll avoid reality for three years or whatever it is. So that's kind of typically been my position. I've started to rethink it a little bit because it does teach critical thinking, at least in the way that we learned it at Chicago, where it was so Socratic. And you know, they just didn't. They didn't care about the bar exam at all. Right. They just assumed that you'd figure it out, which everyone there does. So I really do think my critical thinking skills improved my time in law school. Would you. One, did you. What was your view on whether or not people who didn't want to practice law should go to law school before? And then two, if your view was no has what you just said about critical thinking and AI changed that.
- [31:39]
A
So my view was the same as yours before, particularly if you are not getting into a good enough school where the cost benefit is bad and you're going to end up with a debt load that is a real concern for a lot of people that they take on a ton of debt that better critical thinking couldn't really help.
- [31:58]
B
Well, and a school where their job is to teach you how to pass the bar exam.
- [32:04]
A
Right.
- [32:04]
B
Because if you're occupied this law anyway, what's the point of studying that for three years.
- [32:08]
A
Right. Has AI changed it? Not really. I mean, I had a conversation with a lawyer last week who was like, you know, now that. Now Harvey's telling me how to make my own prompts better, and I'm really starting to wonder how much, you know, how much value I am adding. So I. There are real serious changes coming to the profession that would make me even more hesitant about law as a profession.
- [32:32]
B
As a profession, but what about as the education to provide critical thinking skills?
- [32:36]
A
I think critical thinking education, however you do it, is critically important now because we're now at a point where it just matters.
- [32:45]
B
Like, so you and I have both taught business school. I don't think that that really promotes critical thinking per se. It promotes, like, going on vacations that you call learning trips. So what other professional schools do teach
- [33:00]
A
critical thinking inside that way law school does? Because you have to. Because the whole practice of law is figuring out how to advocate for a position better. Journalism school, maybe. You know, I mean, learning how to be a reporter is 100%, 75% learning how to ask the right questions.
- [33:17]
B
Yeah.
- [33:18]
A
And get information. I think that's actually next time I
- [33:20]
B
talk to a journalist who went to journalism school, I'll ask them what they think about that.
- [33:23]
A
Yeah, but. And then, you know what? This is going to sound stupid, but, like philosophy.
- [33:29]
B
Yeah, I was thinking that too, actually.
- [33:31]
A
I mean, places where they really teach you how to ask hard questions, dissect answers, evaluate the answers that you're getting, and say, does this sound fundamentally right? Based on everything I know. And then ask other questions to test. I. You know.
- [33:45]
B
And what about Divinity school? You spent some time in divinity school. What's with that?
- [33:51]
A
I. I think that one is interesting because you're. You have to ask an answer and work through a lot of hard questions. On the other hand, you're working through a bunch of questions that there are almost by definition, no answers to.
- [34:02]
B
Yeah.
- [34:03]
A
So I, you know, I think that it allows you to better argue for your position, which is not necessarily always critical thinking. Now, the. The stuff I did in divinity school, which was like, highly critical. It was like U of C Law school for, you know, for the practice of religion. It was critical, and that was good. Sure. I gained a lot from it.
- [34:21]
B
And I think Bob and I are both very biased here, but we both went to Ivy League undergrad, and then we both went to Chicago for law school, and then Bob also got his PhD there. I will say that if you were thinking about where to go to school, Especially for graduate school. And you are in a position to go to the University of Chicago for whatever it is you should go. I really do think that one more aside and we'll go back to taxonomy. I was listening to Arthur Brooks's new book and he had one point, because the basic sort of throughput of the book is he teaches at Harvard Business School and his students are these sort of ultimate strivers. And I think we both, having employed HBS graduates, would agree with that. But oftentimes not really thinking critically, more just very linearly trying to achieve the next thing and the next thing and the next thing. And he had this sort of little quiz he gave that where he said, okay, there are five potential things you could think, and they're not mutually exclusive. One is your life is based on the choices that you as an individual make. Two, your life is based on the choices that others make. Three, your life is based on the predetermined laws of nature and physics and biology and everything else. Four, your life is based on the choices and actions of God. And five, your life is based on randomness of the universe. And, you know, I ended up sort of saying yes to almost all of them. A little less on 4 for God. Not because I don't believe in God. I actually believe deeply and pray daily, but I believe in a God that effectively created the universe, but then doesn't really interfere beyond that. But I had, you know, I thought all the others were valid. And I think for me at least, I still think number one's the most because I tend to make unconventional choices. But obviously, you know, the things that we do are still dictated very much by how. What others choose as well. And obviously, you know, no matter what we do, let's say you manage to live to 100 years old, it's still a blip and a blip of a blip of the universe. So that clearly means the laws of nature are highly determinative and randomness is certainly a factor too. How would you answer that question?
- [36:32]
A
So from my standpoint, now I'm going to take a different view on it, which is the question is, what are you trying to get from the question? For me, the question is, how do I gain the most utility in my life from thinking about those answers? And I think ultimately I believe that a. That it's all of them, right? I believe there is randomness. I believe that there are, you know, there are divine or at least periodic or, you know, understandable rules. However articulated, is that number three or number four, that's Number four, probably number three, too. Some of those things are understandable by science. Some of them are not, but that doesn't mean they're not regularities anyway. Some of the things are outside of your control and some things are inside your control. But I think the only one, other than acknowledging all five as being valid and reflecting on that as part of making your life more rich, I think really focusing on number one is the one with the highest utility. I think understanding that you have agency and that your agency and Your reaction to 2, 3, 4 and 5 is going to make your life more or less rich and more or less within the capacity that 2, 3 and 4, or 5 allow. Happy. I think that's the one you can control. And I think you have to take that one most seriously. Just from a utilitarian, or I guess you'd call it pragmatic perspective.
- [37:51]
B
Let me push back a little bit on that because I was talking about this therapy yesterday, and my natural inclination, like as I said, is very much tittling. It's number one. But I think that at least if you have my personality, where you are endlessly applying pressure on yourself to, you know, maximize every second, be productive, create, you know, take on more, and that ultimately becomes detrimental because it's just not healthy. There for me, I think, is some. Some utility in actually accepting the others as a justification to allow myself to do a little less and relax a little more.
- [38:30]
A
Sure. Although you could make the argument that that acceptance is in fact a choice, which is number one.
- [38:35]
B
Oh, yeah. Well, I mean, if you do that, then everything could fit into number one.
- [38:38]
A
Right. That's. I think that's part of the problem, is that, you know, because you are an agent, unless you deny your own agency, number one becomes the one that has the most immediate impact because it's the one. Everything else flows through you. I think from a pure quality of life perspective. I think you have to reach acceptance with all five.
- [38:56]
B
Yeah. Unless someone's truly an atheist and you can't. I don't think their life is worse if they don't accept number four.
- [39:04]
A
I think that's possible. I don't have a way to value it. My personal belief is you have to figure out how to come to terms with all five.
- [39:10]
B
Fair enough. All right, let me let you finish the taxon without interrupting you. And then I have a. Oh, God,
- [39:15]
A
I feel like we're so far past that.
- [39:17]
B
Well, all right, just run through the rest of it quickly.
- [39:20]
A
Okay, quickly. So then the question. Let's say the government can solve catastrophic AI, Right. What is the solution? Then we get to what can we get past stuff that's easy. Is there an easy obvious answer to catastrophic AI? No. If there was an easy obvious answer to not killing everybody, we would have definitely done that already. Right. That's like self explanatory. Is it a crisis? I think actually right now it's not a crisis because despite the fact that everybody can die, the crisis is like a risk benefit situation. And people believe people are, as you know, really bad at assessing low likelihood events. And people, I think in this type of low likelihood event where the risks are tremendously high, underestimate or under or overestimate, but can't reach a consensus on it. So because I don't think there's immediate crisis, I don't think it's necessarily, I think unless they find a way that is easy, it's not passable. And that means you're going to have to find a place where there is a policy consensus, which means, okay, what is something that all of the hyperscaler stakeholders, all of the model stakeholders agree that they can do, that governments agree that they can do and that people will tolerate? And you would think that that was easy to do because the real person you're trying to deal with are the hyperscalers. But the whole hegseth anthropic situation about arming, weaponizing AI was a place where government was the irrational one in terms of the catastrophic risk tolerance, even where the platforms were willing to set a line in the sand. So this is one of these situations where you would think that there was a consensus and then an unlikely stakeholder turned up to have a real issue. You could say the same thing about people who want to say, I know that there's tremendous risk here of catastrophic harm, but my freedom is something that I value and I would like my freedom not to be unhinged, even at the risk of everybody dying. So I right now don't believe that catastrophic harm is something that we've reached a policy consensus on. And I would be surprised in 2027 if we saw landmark legislation on it.
- [41:29]
B
Yeah. Anything else you want to add on that one?
- [41:31]
A
No, I am where we are.
- [41:34]
B
Okay, new topic, which I think it might be even more significant, which you don't even know it's coming. So. But he was already laughing because I previewed it for him, which is, is it truly necessary to wash your hands after you urinate? If you're using a urinal, you don't touch anything and there's no urine on your Hands.
- [41:54]
A
So this is what I. Without going into too much, this, this is what I would call the right hand problem. So let's say you go into the bathroom and you don't touch anything with your right hand, okay? And you can wash just your left hand. Why wouldn't you wash just your left hand?
- [42:09]
B
Right. So then therefore, if it's not like every time you walk into every room, you wash your hand afterwards or after you touch your face.
- [42:17]
A
And like, sure, there are more germs in a bathroom than the average room, but it also gets, it gets cleaned more frequently. So it's not. This is one that is. I think this is much more about other people's perceptions and doing things to make other people feel good that you're not passing germs to them than any real benefit you have to yourself.
- [42:36]
B
Right? Well, okay, but it's perception, right? So, yeah, Hugo said to me, and you can jump on if you want, but like, he's declining. Well, what about the approbation of people if you were. And by the way, I do wash my hands, although I use, I don't use. If I don't pee on my hands, I use water, but not Soap for number one. But Hugo said, well, what about.
- [42:59]
A
That's like 50% of the time.
- [43:02]
B
I don't really recall pinging my hands much since I was like three. But Hugo said, well, wouldn't you worry about someone judging you? And my view is I probably would not. I could see how it's a concern, though. Do you give that much credence?
- [43:19]
A
Well, I need to defend my point
- [43:20]
B
of view on that.
- [43:21]
A
I didn't say. I guess I was asking whether it was the appropr of other people you were concerned about because, like, oh, am I doing?
- [43:29]
B
Am I what Am I putting water over my hands when I don't touch anything and don't have urine on my hands simply in order to avoid.
- [43:35]
A
I guess it's. Why are we even asking the question if, like, what's the consequence of not doing it?
- [43:40]
B
Well, that's okay. Well, that's. So that's where I, where I started the whole thought process was if there's no, if you're not touching anything and there's no urine on your hand, it doesn't seem like there is much consequence in reality for you or anyone else, and therefore why is it necessary? And then the only good argument or somewhat valid argument that I heard, which was from you, was, well, you might not want others to judge you. Now, I don't think I'll stop Doing it simply because it's been 52 years. I'm just used to it at this point, but I think rationally, I'm just not sure. And I guess you could argue that, and maybe the answer is it's not significant. But we're in a global water shortage. It's only growing worse because of extreme climate change. Data centers are using a tremendous amount more water, which further exacerbates the problem. So if you created a new cultural and social norm where men. The idea was men only wash your hands after urinating. If you touched the toilet seat or got urine on your hand, would that save enough water to matter?
- [44:47]
A
I mean, certainly not.
- [44:48]
B
I don't know, Bob, you're the expert on this stuff.
- [44:51]
A
It would matter less than taking a shower that was 50% as long. Right? I mean, you're not talking about a tremendous amount of water, but you are talking about, in aggregate, a lot of flushes. I am going to diverge the topic slightly to throw in something extremely funny that I heard, but it's also sort of on topic, which is, I had this conversation with my daughter about what you do when you tip somebody in the tip jar and the server doesn't see you.
- [45:18]
B
Wasn't there, like a Seinfeld where George, like, tried to make change for himself and got caught? No one saw him put it again, but he got caught stealing from it.
- [45:27]
A
Right. So this is, like, this is the issue. And my point to her is if you are worried about the social opprobrium of others, ultimately, at the end of the day, and people know that you will get manipulated, because people understand if you. If people who are repeat actors with you understand that they can use social norms to. To change your opinion, you will be manipulated over time. Her response to me was, I know all about manipulation. I learned about it in sex ed.
- [45:56]
B
So, Hugo, I don't know if you remember this when the kids were at school, but I remember Abby once, or we got an email. There was something. There was something that made me think that in health they were going to learn sex ed. And then I remember asking Abby about it, and I said, so what are you studying in health? And she said, camping safety. And I'm thinking, like, is that like a metaphor for sex ed? And she doesn't understand it. And I was, like, really confused. And then we were driving somewhere, and she was still, like, relatively young, and she pointed to the side of the road and said, flares. And I was like, okay, they're clearly learning about camping safety.
- [46:30]
A
Safety, yeah. Right.
- [46:32]
B
So I don't know if that helps
- [46:33]
A
you with manipulation or it just suggests to me that people learn a heck of a lot in health care place these days. Much more than I.
- [46:39]
B
Much more than we did. Yeah, but the argument for tipping I think and look one is I've worked for tips. So I think anyone who has worked for tips or has an automatically different perspective on this, but it's just sort of karma for you, right? So like I always over tip unless like the service is truly horrendous. You know, it's partly to be nice to whoever is providing the service, but it's partly the let me put some good in the world. Whatever I'm tipping is not a material amount of money for me and therefore why not redistribute the wealth a little bit?
- [47:11]
A
Right? That was my point, which is if you're tipping, you're not tipping because you want to make a judgment about them. You're tipping because you want to do something slightly good in the world when you can. And whether other people see you tip or don't see you tip, that shouldn't change the fact that that's what your intent is.
- [47:27]
B
I think that's right.
- [47:28]
A
Agreed.
- [47:29]
B
All right, that's a good place to close it. Bob, thanks for joining us.
- [47:31]
A
Thank you.
- [47:34]
B
Firewall is recorded at my bookstore, PNT Netware, located at 180 Orchard street on the lower east side of Manhattan. We'd love to hear from you with questions, feedbacks or idea for a guest. Just email me at Bradleyirewall Media or find me on LinkedIn. And to keep up with what's on my mind and my latest writing, please follow my new substack@bradleytus.substack.com thanks again for listening.