Billionaire Chamath Reveals The "Nash Agent" Productivity Hack

0:00 The spack king Chimoth has an AI agent strategy that he calls his Nash agent to help him make decisions. When a lot of people hear this, they immediately shut down and think, "Whatever, another billionaire rich guy who's going to tell me how he's got agents running his life. Whatever. Another guy smoking the dope trying to make some kind of video suggesting AI is running their life." But that's not exactly what's going on here. In fact, in this video that Chimath put together, the Spack King called If You're Ambitious But Inconsistent, please watch this. Chimath gives a few components of advice. One, he talks about certain apps, which we'll talk about. He talks about maxing, and then he talks about the Nash agent, this AI agent he uses to help him make decisions in life. And what I found really useful was not his Nash agent, which was really complicated and somebody who's taken classes on game theory before. I get it and I'll explain it. But we're going to take everything together, simplify it, tone down the stress, and try to actually make some money with this from the point of view of an operator. How can you actually operate with the advice that Chimamath just provided and turn it into money?

1:21 So, first things first, I want to be really clear about this. Chimamoth was appropriately transparent that AI is like a good starting point. It's still not the best at making all decisions in the best possible manner. But what he argues is that he uses a nosh agent to gify decisions that he makes and that gives him a few advantages. We're going to talk about those advantages. But first, when you hear this nosh phrase, you might get a little overwhelmed or confused. Don't worry, I'm going to make this really simple with an example. So, let's think about this for a moment.

2:00 Let's think about a scenario where you have a girlfriend and a boyfriend who are stuck at work and both of their phones are dead and they don't have phone chargers. Okay? So, girlfriend's stuck at work, dude stuck at work. Both of them have the option to go home right now and hang out with their girlfriend or boyfriend. The problem is they don't know if the other person is also going to make the decision. Best case scenario, both of them go home and then they get to work out together or hang out together, right? This is called the NSH equilibrium. This is an example where the highest expected value is provided to each side.

2:43 Each person goes home, hangs out together. Okay? But in life, we can't always assume that we're always going to get the best solution, the most optimal solution, which in this case is both of them magically decide to go home. The second best potential scenario is that they both decide to stay late at work. Even though they don't really want to work as much as they'd like to hang out, so their net value is a lower number, right? Their utility is a lower number.

3:11 They at least both get to make a little bit of extra money and they didn't go home and the other person wasn't there and then they got burned, right? That would be the ultimate worst case scenario. See, in this case, I would actually argue that the total utility is probably lower than 2 + 1 because the person who goes home probably ends up even more disappointed at one in this case. and the person at work probably ends up feeling guilty that they didn't go home and so their two is probably even lower. Now, I don't want you to over complicate this. The point of this is there are two scenarios here that work well. One is better than the other.

3:48 This is the best Nash equilibrium. But this is another also good option. And the point of what Chimamath is trying to say is he uses AI not to tell him which of these four options is the absolute best to do. But what he's going to do is, and this was my reading of it, right? What he's going to do is he's going to use AI to talk to an agent to say, you know what, let's at least eliminate the bad options, right? That way, even though we're now choosing between two options where one is more ideal than the other, assuming we have the choice beyond just randomness, which in this case you have a randomness element, right? If people's phones turn on, they communicate, that randomness goes away. But in this example here, Chimoth by calling his agent, the Nosh agent, he's really trying to say, I want to slow down. I want to see my blind spots more clearly. I want an agent to tell me, look, don't feel like you're in the cereal box aisle where you have all of these choices. Let's just tell you bluntly, this is bad because there are some really big holes here. This is also bad because there are some really big holes here. And now let's slow down how you negotiate, how you put together terms of a contract or how you decide different directions between products, which is what Chimamoth talks about. And then you can sit, you know, in Chimant's example, he says he ruminates over this with his wife and was like, "Okay, well, what's going to be the most optimal scenario, whatever, at least you've simplified the decision-m process." And so, I always like to think of myself as an operator, somebody who just gets things done. And I'm not perfect, so I'm not trying to pat myself on the back. I just like getting things done.

5:38 And I have to say, I've also used artificial intelligence in this manner where it's really, how should I put it? It's really seductive to run everything by your AI and then kind of sit there and be like, "Yeah, yeah, that's right. Option three and four suck. Option one or two are the best." But then there's a problem, and this is what I find is really, really common. A lot of people already know they could use AI to help them concentrate a decision. And this is the part that Shamath didn't talk about in his video, at least not when it came to the AI agent. I think he could have stressed this a little bit more is, and this isn't to dish on Shimath, it's just to say once you talk or come up with sort of like the best two or three options, whatever it is in your case, you have to choose one. Because if you don't choose one and then you keep dwelling or you ingest that into other AI agents, now you're effectively demanding that an AI agent that's good enough to give you a start be good enough to do it all for you. And this getting you a start, giving you the right direction up front, you could do that. That's reasonable. Expecting AI to do everything for you and then just ultimately not doing anything. That's the big failure point. So, I see a plus and a minus here that Shmath talks about. Yes, get slowed down. Remove the emotion from your decision-m process. As a real example, let's say somebody, and it happens to everybody at some point in the future, well, maybe not everybody, but it happens to a lot of people, especially business owners or entrepreneurs. Let's say one day you get sued and you're like, "This is a scam.

7:24 This is a total fraud. This is a ripoff. we will cream these people in court and let's pull up an expected value chart for example. So let's go back to the Nash equilibrium over here and let's just say hey we've got the emotional point of view. So we'll just sort of resquare this out right here. We've got the emotional point of view. and it says we're going to go to, this is going to be settle. and then we have jury is going to be the, next option, which, you know, that could take you a couple years to get to, but, let's call it a jury over here. So, you've got settle and you've got jury. And then what we'll do over here is we'll say win and then we'll say lose.

8:12 Sometimes what an AI agent can help you do is sort of redefine what a win actually is and remove some of the emotion. Because usually what happens in legal battles is people are convinced that they have to go to a jury and that the jury is going to like them. If a jury doesn't like you, you're screwed. So in this case, if you go to a jury and win, okay, that's best case scenario for you. it would seem, but it probably cost you a lot of money. And are you going to be able to collect money back from the person that you won against? Maybe not necessarily. So maybe the you know, jury win is like a good ego win, but maybe it's only a three in terms of utility. maybe settling early for a number that is really low and you don't have to blow all this money on attorneys and the time and the effort or whatever is actually the best win. So, we'll call this a early settle and then this would be like a late settle which would be pretty bad and then of course going all the way paying all your attorney fees and then paying the other person's attorney fees and losing at jury would obviously be the absolute worst case scenario, right? So, in this case, there are two equilibria. Well, I mean, I'm not putting the second person in here, obviously, but for your decision-m tree.

9:39 There's a best case scenario where you just settle out early and win. The other person goes away. You negotiate it away initially or you win at jury, but the reality is this one's going to have more utility for you because you'll save years of time. You'll save the emotion. And even though you'll have to bite your tongue over here, the fact that somebody could be ripping you off and then you're in sort of settle mode, at least it's over and you can focus on what you're actually good at. And so, in my opinion, I pick up what Chimoth is trying to throw down, trying to say, look, we could ideulate with an AI, which is pretty good at helping us identify some blind spots, but it's not perfect. It's a starting point to help us think like a computer. We're going to think a little bit more formulaically. And I think that's what he's getting at here is we're going to think formulaically rather than just throwing emotion into everything. and then we can make a real decision. I like that. Now, the next parts of his video that he brought up were things that are really valuable, but I think are somewhat less useful than thinking about everything from a starting point as an AI agent. So, start as an AI agent. Let's take the emotion out. Then this is where the best order number two in my opinion comes in. His argument of maxing is don't overthink a new job or career idea or license or education. Just do the thing and then if you don't like it, stop. I actually think I'm somebody who has maxed. I not only became a licensed real estate agent, real estate broker, real estate lender twice, licensed drone pilot back in like 2016.

11:29 That was actually a big deal. then I got my series 65 financial advisory license. I passed the test for the series 7, the 24, the 27, the 14, the 63. I mean, there's so many financial advisory licenses and financial regulatory licenses that have passed. there was even a part where I'm like, "Oh, I want to I want to be a lawyer." I didn't end up going down the bar direction. and instead I became a pilot and I'm like, I'm not just going to be a regular private pilot. I'm going to be instrument rated because I want to fly in the clouds. And you know what? I want two engines. So I'm going to become a multi-engine rated pilot. And you know what? Then I'm going to fly a jet. I'm going to fly my own jet. And I did that, you know. So, the very cool thing about maxing that Shamath argues here is if you could bring yourself to just do it. You know what? I want to be a pilot. Okay, go go go take your first flight lesson. Okay, I want to be a real estate agent. How hard is it to go online and sign up, get real estate license in California, pay the 300 bucks or whatever, and start taking the courses. Boom, you'll get your license within like 6 months. And if you like it, keep doing it. And if you don't, don't do it anymore. I didn't even think I mentioned that. At one point, I was running a licensed contracting company, and I'm like, man, I make way more money selling houses than doing licensed contracting. So, I just flip-flopped on the contracting.

12:52 Now, I don't regret that. I actually look back and say, "Hey, all those skills are really useful today because I know exactly what not to do and I know exactly where I can make money with my time and make more money with my time. If I don't like something, just stop." Love that advice from Chimoth. Really valuable. Now, the last set of advice Chimoth gives, I think at this point has become generic. It's been something that I've been talking about for a very long time. I even talk about it in sort of my daily wealth blog that I'll send out which you could get for free in the Meet Kevin app. You just download Meet Kevin your in the Google or Android app store and he talks about eliminating doom scrolling. Okay, this is at this point cliche. He actually suggests hey have a second phone and only use that for Tik Tok or Instagram or whatever. And I find it really interesting. This is coming from the social media guy Chimath, you know, one of the Facebook guys.

13:45 But he's kind of got a good point. You waste a lot of time vertically scrolling and I don't allow myself to vertically scroll unless I'm on vacation. And I do that to remind myself how useless it is to doom scroll. So if on I'm on vacation I'm allowed to look. And I might spend 3 hours over the course of 2 or 3 days where it's like, wow, I spent an hour there, 45 minutes there, an hour 15 there or whatever. And I'll even scroll sometimes with my kids and just sort of look around. And then I'll look the next day and I'll be like, "Wow, from the three hours that went into that while we were sitting on the beach in Hawaii or, you know, on the patio in Hawaii, what did I take away?" Not much. You'd be better off watching like an HBO comedy special or something. It's literally a timekilling. but worse, Chimath actually says that doom scrolling will eventually make you feel like you're inadequate because you're comparing yourself to all these people's superficial lifestyles.

14:42 All right, little little extreme. I look at it as more of a productivity time suck or a sleep suck. Like if you're keeping yourself up at night and then you sleep less and then you're tired for work, that's pretty bad, too. So, from a practical point of view, I agree with Chimoth here. Very last thing he kind of sprinkled in there was this idea about stop ruminating. I love this. A lot of people spend a lot of time thinking about their regrets and because they think so much about their regrets, they spend less time doing. Just get out there and do. So overall really positive message here from Shmath and look forward to seeing some more.

15:16 >> Why not advertise these things that you told us here? I feel like nobody else knows about this. >> We'll we'll try a little advertising and see how it goes. >> Congratulations, man. You have done so much. People love you. People look up to you. >> Kevin Praath there, financial analyst and YouTuber. Meet Kevin. Always great to get your take.