WEBVTT - Should You Learn Poker from ChatGPT? And Other AI Questions 0:00:15.564 --> 0:00:28.084 Pushkin. Welcome back to Risky Business, our show about making 0:00:28.124 --> 0:00:30.564 better decisions. I'm Maria Kanikova. 0:00:30.484 --> 0:00:31.444 And I'm Nate Silver. 0:00:32.804 --> 0:00:35.204 Today on the show, we're going to be getting into 0:00:35.284 --> 0:00:38.284 it on AI. There has been a lot of news 0:00:38.324 --> 0:00:40.844 on the AI front in the last few weeks, some 0:00:41.004 --> 0:00:44.244 of it coming from the Trump administration and some coming 0:00:44.284 --> 0:00:47.964 from overseas with China and Deep Seek. 0:00:48.244 --> 0:00:51.284 Maria thinks AI is overrated. I think AI is properly rated, 0:00:51.804 --> 0:00:52.884 as you'll see. 0:00:54.204 --> 0:00:56.564 And then we're going to get into a listener question 0:00:56.684 --> 0:01:01.404 about poker and how to beat your local cash can. 0:01:15.684 --> 0:01:19.924 Let's start with our friend artificial intelligence. 0:01:20.204 --> 0:01:23.204 Yeah, so I think we're talking about a few different things, right, 0:01:23.564 --> 0:01:26.044 So we have the Trump initiative on AI. So we 0:01:26.124 --> 0:01:31.804 have a few things that he did, including things that 0:01:31.884 --> 0:01:36.404 he took away, right the executive Order that took away 0:01:36.484 --> 0:01:39.044 certain restrictions on AI that had been put in place 0:01:39.844 --> 0:01:43.524 by Biden. But then also we have you know, this 0:01:43.684 --> 0:01:48.524 big funding initiative into AA by the US government, Stargate, 0:01:48.644 --> 0:01:51.724 and then we also have AI coming out of China 0:01:52.164 --> 0:01:55.044 Deep Seek, which has freaked everyone the fuck out. I 0:01:55.044 --> 0:01:58.764 think that's the scientific way of putting it and has 0:01:58.804 --> 0:02:03.644 made markets, crash video stocks, a lot of other stocks Nasdaq. 0:02:03.924 --> 0:02:07.764 You know, people have not been happy to see the 0:02:07.804 --> 0:02:10.844 success of deep Seek, So there's a lot to talk 0:02:10.844 --> 0:02:11.404 about today. 0:02:12.644 --> 0:02:14.964 Where do you want to start, Maria. 0:02:15.084 --> 0:02:17.804 Well, do we want to start with? I think if 0:02:17.884 --> 0:02:21.204 it makes sense to start with deep seek because it actually, 0:02:21.844 --> 0:02:24.284 you know, I think that the two are very related, 0:02:24.604 --> 0:02:27.244 right because we have what the US government is and 0:02:27.444 --> 0:02:29.924 is not doing, and one of the reasons that right 0:02:29.924 --> 0:02:33.364 now things are freaking out is because of deep Seek. 0:02:33.404 --> 0:02:35.644 So I think we can start with that, see what 0:02:35.684 --> 0:02:37.804 the implications are, what the responses can be, and then 0:02:38.004 --> 0:02:40.484 kind of see what the US has done so far 0:02:40.604 --> 0:02:42.724 to see if it's in line with what we think 0:02:42.924 --> 0:02:44.404 the correct strategy should be. Yeah. 0:02:44.404 --> 0:02:48.604 Look, until about a week and a half ago, when 0:02:48.644 --> 0:02:52.204 people started to notice Deep Seek their model are one. 0:02:52.244 --> 0:02:56.124 In particular, the conventional wisdom was that, like America is 0:02:56.244 --> 0:02:59.204 way ahead in the AI race. I should say with 0:02:59.284 --> 0:03:04.284 respect to large language models, machine learning transformers, right, you know, 0:03:04.404 --> 0:03:07.444 driverless cars is a different enterprise, and drones and things 0:03:07.484 --> 0:03:11.204 like that. But in terms some LM's large language models 0:03:11.244 --> 0:03:16.004 chatcypt like things. Then you know, us probably one, two, three, 0:03:16.084 --> 0:03:19.884 four in the rankings and so this, Yeah, this has 0:03:20.004 --> 0:03:23.764 interesting geopolitical implications. Yeah. 0:03:23.804 --> 0:03:26.564 And the one of the reasons, just to step back, 0:03:27.164 --> 0:03:30.164 that the US was assumed to be ahead was because 0:03:30.404 --> 0:03:33.004 the United States had made it more difficult for foreign 0:03:33.044 --> 0:03:37.564 governments to acquire like China, especially to acquire the chips 0:03:37.644 --> 0:03:43.044 that are necessary to build these large AI models, the resources, 0:03:43.044 --> 0:03:45.964 et cetera, et cetera. And so one of the disconcerting 0:03:46.004 --> 0:03:49.524 things to the United States was that when Deepseak announced 0:03:49.524 --> 0:03:52.044 as results, it also announced that it basically was able 0:03:52.084 --> 0:03:55.484 to do this at one tenth of the cost and 0:03:55.604 --> 0:03:58.764 resources that other models had used. So this was like 0:03:58.804 --> 0:04:01.804 an oh shit, you know, even if we restrict access 0:04:01.804 --> 0:04:04.724 to all of these other things, they're still able to 0:04:04.724 --> 0:04:07.644 do this. Now, I will say, and other people have 0:04:07.724 --> 0:04:10.764 pointed this out, I don't actually know how much we 0:04:10.804 --> 0:04:13.204 can trust the numbers and figures, right, this is just 0:04:13.324 --> 0:04:16.044 a claim. We don't know what the training materials were, 0:04:16.244 --> 0:04:18.644 we don't know what the development costs actually were. We 0:04:18.724 --> 0:04:21.204 just know what they claim that they were. So I 0:04:21.244 --> 0:04:23.164 think that this is something that we should put an 0:04:23.204 --> 0:04:26.524 asterisks next to because it is important to realize, right 0:04:26.964 --> 0:04:30.724 that if you can't if you can't actually brace the 0:04:30.804 --> 0:04:34.444 information and verify it, and you have to take it 0:04:35.004 --> 0:04:38.244 on faith. That's never a good way to take information. Right, 0:04:38.244 --> 0:04:40.364 That's one of the things we say over and over 0:04:40.604 --> 0:04:43.644 on risky business when you make decisions, try not to 0:04:43.684 --> 0:04:46.324 take things on faith. Try to verify, right, That's that's 0:04:46.404 --> 0:04:47.924 much better. Yeah. 0:04:47.964 --> 0:04:53.524 So this is basically built by a Chinese hedge fund, right, 0:04:53.564 --> 0:04:57.484 which is well which is well capitalized. And you know, 0:04:57.564 --> 0:05:00.204 there are a lot of smart computer engineers in China, 0:05:00.284 --> 0:05:02.164 and I think they were all working on this project. 0:05:02.204 --> 0:05:05.084 So so kind of the last step of training. I mean, 0:05:05.124 --> 0:05:08.364 it's it's a little bit like you know Rosie who 0:05:08.484 --> 0:05:09.244 Rosie Ruiz? 0:05:09.844 --> 0:05:11.324 I do know who Rosie is is? 0:05:11.764 --> 0:05:14.084 Yeah, if you've covered like fraudsters, right, And I don't 0:05:14.084 --> 0:05:15.764 mean to say it's a fraud, but it's like, so 0:05:15.884 --> 0:05:18.044 she runs what was the New York Marathon? Right, who? 0:05:18.164 --> 0:05:19.644 Like it was a fraud. 0:05:19.684 --> 0:05:22.604 By the way, So Rosie Ruiz actually figures in my 0:05:22.684 --> 0:05:29.124 next book on cheating. So Rosie Ruiz one quote unquote 0:05:29.124 --> 0:05:33.604 I'm putting this in quotes. The Boston Marathon women's time 0:05:33.884 --> 0:05:38.724 with this incredible story was amazing, wonderful. You know, people 0:05:38.844 --> 0:05:40.684 loved it, and then it turned out that she took 0:05:40.724 --> 0:05:44.164 the subway for a huge portion of the race, and 0:05:44.804 --> 0:05:46.764 but she almost got away with it, which is the 0:05:47.244 --> 0:05:50.404 really fucked up thing. The reason she got caught was 0:05:50.404 --> 0:05:52.844 because the subway car that she happened to get on 0:05:53.324 --> 0:05:56.044 had a reporter who was a photographer who had been 0:05:56.084 --> 0:06:00.804 covering the marat on and was like, wait, what's going on. 0:06:01.284 --> 0:06:04.604 So it actually took them multiple days to figure out 0:06:04.684 --> 0:06:07.964 that Rosy Ruiz did not actually win, did not actually 0:06:08.044 --> 0:06:10.564 run the time that she shran. By the way, this 0:06:10.644 --> 0:06:14.164 was back in nineteen eighty so right the technology was different. 0:06:14.644 --> 0:06:17.764 People were not getting tracked as closely as they are 0:06:17.844 --> 0:06:18.244 right now. 0:06:18.444 --> 0:06:19.884 But spoiler alert for. 0:06:20.044 --> 0:06:24.684 You know, or big fun thing that I get into 0:06:24.724 --> 0:06:27.164 in my book, it's possible even today to pull a 0:06:27.284 --> 0:06:27.964 Rosie release. 0:06:28.404 --> 0:06:30.724 So that's for later. But Nate, why are we talking 0:06:30.724 --> 0:06:31.964 about Rosie. 0:06:32.004 --> 0:06:35.324 Because it's a little bit like if I don't say 0:06:35.324 --> 0:06:38.364 there's any actual fraud, I really don't trust anything coming 0:06:38.404 --> 0:06:42.284 out of mainland China. Yeah, but like but yeah, so 0:06:42.324 --> 0:06:44.484 it's a little bit like if you run and around 0:06:44.524 --> 0:06:47.444 the twenty fourth mile and then run two really good 0:06:47.484 --> 0:06:49.564 closing miles, it's still not the same, and it's like 0:06:49.644 --> 0:06:52.284 not like an accurate representation to say that just the 0:06:52.284 --> 0:06:54.444 cost of this training run when you have all these 0:06:54.484 --> 0:06:57.084 resources behind it, and it's kind of like the last step. 0:06:58.484 --> 0:07:01.884 But clearly it's more efficient in terms of you know, 0:07:01.924 --> 0:07:06.964 when you run a request, put in a query, how 0:07:07.004 --> 0:07:09.204 many computer cycles is a burning through? I'm using very 0:07:09.244 --> 0:07:14.244 nigh technical language here, right, You can actually host deep seek. 0:07:14.844 --> 0:07:16.404 You keep going to call deep stack. There are a 0:07:16.404 --> 0:07:19.844 lot of deep stack poker terms. You can host deep 0:07:19.884 --> 0:07:24.004 seek at on a desktop. Right. If you want to 0:07:24.004 --> 0:07:26.604 actually have it say things about Tanaman Square, for example, 0:07:26.684 --> 0:07:29.084 then you need to run your native instance because the 0:07:29.124 --> 0:07:32.004 official Chinese hosted web version doesn't like to you know, 0:07:32.044 --> 0:07:33.684 doesn't like to talk about certain things. 0:07:33.964 --> 0:07:36.444 I would say, Maria, you don't say that, you don't 0:07:36.484 --> 0:07:37.724 say so. 0:07:38.084 --> 0:07:39.964 Then there are a lot of debates though about what's 0:07:40.004 --> 0:07:42.844 it mean if it turns out that I mean so. 0:07:42.884 --> 0:07:45.244 One category to bat is like does the US lead 0:07:45.724 --> 0:07:49.084 lose its lead versus China. That's one kind of whole bucket, right. 0:07:49.644 --> 0:07:51.884 Another bucket is like, what's it mean if you now, 0:07:51.964 --> 0:07:55.044 like run an AI lab on a desktop and they're 0:07:55.044 --> 0:07:56.764 only going to get faster how do we regulate this? 0:07:56.804 --> 0:07:58.804 And then there's a whole bunch of economic stuff about, 0:07:58.844 --> 0:08:02.484 like you know, what's this mean for the price of 0:08:03.204 --> 0:08:04.124 different assets? 0:08:04.204 --> 0:08:04.364 Right? 0:08:04.444 --> 0:08:09.284 So Nvidiah, for example, is the largest manufacturer of semiconductor, 0:08:09.644 --> 0:08:15.604 is a Taiwanese company, you know, so if compute is cheaper, 0:08:15.724 --> 0:08:17.804 is that good for them or bad for them? It's 0:08:17.804 --> 0:08:20.364 not necessarily straightforward. 0:08:19.764 --> 0:08:24.004 Right, Yeah, it's absolutely not straightforward. And also, you know, 0:08:24.124 --> 0:08:26.964 one of the one of the other potential things that 0:08:27.044 --> 0:08:30.004 happened with deep seek we don't know is a process 0:08:30.244 --> 0:08:33.084 of training known as distillations. So that's a slightly more 0:08:33.404 --> 0:08:36.724 technical term, but what it actually means is that you 0:08:36.884 --> 0:08:40.244 are training your model on the outputs of other models, right, 0:08:40.444 --> 0:08:43.164 so you can actually pass the benchmarks and be able 0:08:43.204 --> 0:08:46.004 to trade up more quickly because you are using Okay, so. 0:08:45.964 --> 0:08:49.444 Maybe now we are having more of a Rosie Ruize situation. 0:08:49.004 --> 0:08:51.004 And if that happens, then then it is more of 0:08:51.004 --> 0:08:54.564 a Rosie Rueze situation and just as a like as 0:08:54.564 --> 0:08:57.404 a flag, that's not legal, right, Technically, you're not supposed 0:08:57.404 --> 0:08:59.924 to do that. But if you do do it and 0:08:59.964 --> 0:09:01.284 you're able to then cover. 0:09:01.204 --> 0:09:05.284 These outputs, the respect for intellectual property varies by countries. 0:09:05.324 --> 0:09:07.484 I don't want to be I don't want to bad 0:09:07.564 --> 0:09:10.604 mouth the culture for not respecting electric property rights. But yeah, 0:09:10.604 --> 0:09:12.884 if you're if you're you know, if you can just 0:09:12.924 --> 0:09:16.644 copy off chat, GPT or in for what its weights are, 0:09:16.844 --> 0:09:19.124 then that you know, I mean, that's that's a that's 0:09:19.124 --> 0:09:19.644 an issue. 0:09:20.564 --> 0:09:24.164 Yeah, it absolutely is. But let's let's flip that actually 0:09:24.204 --> 0:09:27.044 around a little bit to talk about some of the 0:09:27.044 --> 0:09:32.124 potential positives of this, which you know there we're obviously 0:09:32.124 --> 0:09:36.284 seeing you know, potential uh red flags and negatives. 0:09:36.764 --> 0:09:38.244 But what about I mean. 0:09:38.364 --> 0:09:41.524 I actually think that it's not a bad thing that 0:09:41.604 --> 0:09:45.044 deep seek is open source, right, so people can try 0:09:45.044 --> 0:09:47.604 to figure out, Okay, what is it actually doing, how 0:09:47.684 --> 0:09:48.564 is it actually doing? 0:09:48.604 --> 0:09:50.844 It? Isn't that? Isn't that good? 0:09:51.124 --> 0:09:52.684 And I know that, and I know that one of 0:09:52.724 --> 0:09:55.804 the things that you know that people don't often like 0:09:55.924 --> 0:09:59.044 is open source. But there's like open source can go 0:09:59.164 --> 0:10:02.524 either way, right, open source bad if it's US and 0:10:02.604 --> 0:10:05.924 China's getting its secrets, but you know, open source good 0:10:05.964 --> 0:10:09.044 in other respects. And I think open source is one 0:10:09.084 --> 0:10:11.764 of the few ways that we can actually peek inside 0:10:11.764 --> 0:10:14.244 the black box. And I would be much more concerned 0:10:14.284 --> 0:10:16.324 if it wasn't even open source, right, if we had 0:10:16.324 --> 0:10:17.844 no idea how it was trained, if we didn't know 0:10:17.844 --> 0:10:19.444 about the funding, if we didn't know it of this 0:10:19.764 --> 0:10:22.684 and it was an open source right, if all of 0:10:22.684 --> 0:10:26.604 those things were compelled, I'd be more worried. Since it 0:10:26.724 --> 0:10:29.204 is open source, I actually think that could potentially be 0:10:29.284 --> 0:10:31.964 a good thing in terms of knowledge sharing and trying 0:10:32.004 --> 0:10:34.964 to figure out how do we make these processes more efficient? 0:10:35.004 --> 0:10:38.244 And Meta has actually created a workforce that is studying 0:10:38.284 --> 0:10:41.084 the deep sea processes and figuring out, okay, how can 0:10:41.124 --> 0:10:43.444 we use these and do we want to change the 0:10:43.484 --> 0:10:47.364 way that we're developing some of our llms, some of 0:10:47.404 --> 0:10:50.684 our chat you know, models to mimic their processes so 0:10:50.724 --> 0:10:55.924 that we become more efficient. And if this process actually 0:10:55.964 --> 0:10:59.324 means that we use fewer resources, that's obviously a net 0:10:59.364 --> 0:11:03.004 positive for the environment. However, if it means proliferation of 0:11:03.044 --> 0:11:05.684 ais everywhere and it's cheaper than could actually be a 0:11:05.724 --> 0:11:08.324 net negative. Which is all to say that this is 0:11:08.364 --> 0:11:11.684 really complicated. This is not black and white, and when 0:11:11.684 --> 0:11:13.804 you're making these sources of decisions, you have to assign 0:11:13.804 --> 0:11:16.084 weights to all of these different outcomes, all of these 0:11:16.124 --> 0:11:19.124 different probabilities, and frankly, I don't have the I don't 0:11:19.124 --> 0:11:21.444 think anyone has the expertise to do that because it's 0:11:21.484 --> 0:11:23.884 such a new world. I certainly don't have the expertise, 0:11:23.924 --> 0:11:26.844 but I don't think anyone can see the future clearly 0:11:26.964 --> 0:11:28.724 enough to figure out, you know, how do we wait this? 0:11:28.844 --> 0:11:31.764 Because there's a lot of uncertainty around this. 0:11:32.084 --> 0:11:38.564 So the hardcore AI concern people, the doomers, tend to 0:11:38.564 --> 0:11:42.444 be anti open source. Open AI was founded as open AI, 0:11:42.684 --> 0:11:45.764 but no longer kind of abides by that mission, and 0:11:45.804 --> 0:11:49.764 I think so I think their view. Let me try 0:11:49.764 --> 0:11:53.644 to give you the summary slash kind of Steelman version 0:11:53.684 --> 0:11:59.204 of it. Right, these think, okay, this period where AI 0:11:59.404 --> 0:12:02.844 is being birth, where artificial general intelligence ranging up to 0:12:02.964 --> 0:12:07.204 artificial superintelligence. The first means it can do most things 0:12:07.244 --> 0:12:10.364 at a human level or human life well plus right, 0:12:10.484 --> 0:12:13.404 and super intelligence means it achieves breakthroughs that no human 0:12:13.524 --> 0:12:17.804 can across the broad range of fields. Basically, right, it 0:12:17.884 --> 0:12:20.444 thinks this process of giving birth to these models is 0:12:20.484 --> 0:12:24.044 going to be very dangerous, and therefore you want to 0:12:24.084 --> 0:12:26.324 have it in the hands of as few people as 0:12:26.324 --> 0:12:29.644 possible and trusted people as possible. I think at one 0:12:29.644 --> 0:12:32.724 point Sam Malton was trusted by this community. You know, 0:12:32.804 --> 0:12:36.324 Anthropic is a bunch of people who left open AI 0:12:36.444 --> 0:12:38.364 because they thought open a I was moving too fast. 0:12:38.444 --> 0:12:42.844 Right and then and then Google Gemini. You know, Google 0:12:42.844 --> 0:12:47.084 again had been fairly conservative about how it moved on AI. 0:12:47.284 --> 0:12:49.444 Didn't want to you know, it's a huge enterprise, doesn't 0:12:49.444 --> 0:12:51.244 want to risk that and other things and kind of 0:12:51.444 --> 0:12:56.164 a different culture I would say at Google, then Facebook 0:12:56.204 --> 0:13:00.684 or Meta, the reason why their models were open source 0:13:00.844 --> 0:13:04.324 is because, like they weren't competitive. This is what people 0:13:04.364 --> 0:13:07.004 would say in the AIXPEC community, right, is because they 0:13:07.044 --> 0:13:11.124 weren't as good as these, you know, clearly as Anthropic 0:13:11.124 --> 0:13:13.484 which has clawed, or open ai or clearly the two 0:13:13.524 --> 0:13:16.964 leading ones. And then and then you know, Google Gemini 0:13:17.084 --> 0:13:20.404 had some problem with like drawing like woke Nazis and 0:13:20.404 --> 0:13:23.364 stuff like that, which I you know, I think it's 0:13:23.404 --> 0:13:25.804 not as good as the other two by a fair bit. 0:13:25.844 --> 0:13:27.524 But like, but that was that was considered in the 0:13:27.524 --> 0:13:31.484 pecking order, right that Microsoft was excuse me, that Meta 0:13:31.844 --> 0:13:35.244 too many m's was fourth, and therefore they open source 0:13:35.284 --> 0:13:37.124 it to kind of say, okay, well we're gonna like 0:13:37.444 --> 0:13:41.684 in some sense it's it's uh, not quite like sabotage. 0:13:42.124 --> 0:13:43.564 You're like, well, fuck you, You're not going to get 0:13:43.564 --> 0:13:46.484 these same returns to us, and will have applications people 0:13:46.484 --> 0:13:48.404 who want to pay for it for free or whatever. Right, 0:13:49.364 --> 0:13:52.524 So the fact that it's open source is interesting. I mean, 0:13:52.524 --> 0:13:58.324 you know, China clearly is willing to do things, uh 0:13:58.724 --> 0:14:05.004 to undermine the American even if so. I mean, the 0:14:05.004 --> 0:14:07.764 other big one in the news is of course TikTok, 0:14:07.764 --> 0:14:10.444 which is owned by bike Dance. Right, the US Congress 0:14:10.444 --> 0:14:13.004 passes a law upheld by the courts. It says you 0:14:13.084 --> 0:14:17.564 have to sell to an American company or at least 0:14:17.644 --> 0:14:19.404 a country not listed on like the in there's like 0:14:19.404 --> 0:14:22.204 literally like an enemy's list of countries, and China is 0:14:22.284 --> 0:14:27.644 on it, right, And bike Dan says, we have this 0:14:27.684 --> 0:14:29.604 really valuable enterprise, but we'll just turn it off, right, 0:14:29.684 --> 0:14:32.284 If you make us sell it, then we'll just turn off. Well, clearly, 0:14:32.324 --> 0:14:34.284 I mean obviously it would be a forced sale and 0:14:34.324 --> 0:14:36.644 you wouldn't have quite the market clearing price, but like 0:14:36.924 --> 0:14:39.124 you know, the value is a lot more than zero, 0:14:39.284 --> 0:14:41.004 and they're willing to So you know, there's a little 0:14:41.004 --> 0:14:43.524 bit of suspicion that like this is just meant to 0:14:43.724 --> 0:14:47.964 like undermine America's lead in AI and not make a 0:14:47.964 --> 0:14:50.124 whole lot of profit for China. That's kind of like 0:14:50.164 --> 0:14:53.204 addition by subtraction, right, you know, And people say, hey, 0:14:53.204 --> 0:14:56.364 these people are idealistic. It's not the same capitalist system there. 0:14:56.404 --> 0:15:00.564 And even in the US sometimes the founder aren't that greedy. 0:15:00.564 --> 0:15:02.204 They just want to make a really cool product. So 0:15:02.644 --> 0:15:04.444 there might be some of that too, But it's in 0:15:04.484 --> 0:15:07.724 line with previous Chinese strategy, I suppose. 0:15:08.484 --> 0:15:11.404 Yeah, it's actually interesting that you that you mentioned TikTok 0:15:11.444 --> 0:15:14.804 because this is another kind of strategic element of this 0:15:15.084 --> 0:15:19.964 that you know, we're we being the US has moved 0:15:20.004 --> 0:15:21.204 to band TikTok right now? 0:15:23.044 --> 0:15:26.764 Sorry, what USA? 0:15:27.204 --> 0:15:33.644 You are mooting for the US? Nice and nice? All right? 0:15:33.764 --> 0:15:38.804 Tem U s I is uh was moving to band TikTok. 0:15:38.844 --> 0:15:40.404 We have no idea what's going to happen with that now. 0:15:40.604 --> 0:15:43.444 But in some ways, you know, and deep seek is 0:15:43.564 --> 0:15:46.724 just like la la la la la. Right, There's there's 0:15:46.764 --> 0:15:49.844 no movement in that direction, and instead Trump is talking 0:15:49.884 --> 0:15:53.644 about tariffs and other things and and trying to kind 0:15:53.644 --> 0:15:55.364 of get at it that way. And I think that 0:15:55.444 --> 0:15:58.084 it's a very interesting dichotomy where like, if you're worried 0:15:58.084 --> 0:16:01.044 about TikTok, like, shouldn't you be worried about the strategic 0:16:01.404 --> 0:16:05.044 and security risks of you know, of a company that 0:16:06.084 --> 0:16:08.844 runs you know that all of the idol Ai models 0:16:08.884 --> 0:16:12.044 are run into right, Like that's Chinese owned, Chinese developed, 0:16:12.324 --> 0:16:15.524 Like do you if you're going to be consistent like that? 0:16:15.524 --> 0:16:17.724 That seems to be a much greater risk than TikTok 0:16:17.804 --> 0:16:18.764 to be perfectly honest. 0:16:21.644 --> 0:16:23.524 And we'll be right back after this break. 0:16:33.204 --> 0:16:36.844 So what do we, like, what can we take from 0:16:36.884 --> 0:16:39.644 this and from what's happened in the last week. How 0:16:39.684 --> 0:16:45.364 does that mesh with the types of endeavors that that 0:16:46.284 --> 0:16:50.004 Trump and team have already put forward. Now, one of 0:16:50.044 --> 0:16:52.404 the things, so I started off by saying that they've 0:16:52.444 --> 0:16:56.844 rescented the executive order. The executive order that Biden had 0:16:57.284 --> 0:16:59.884 on Ai did have to do with open source, right, 0:17:00.164 --> 0:17:03.124 and it did have to say it was actually very 0:17:03.124 --> 0:17:06.724 skeptical of open source as well, because it wanted, you know, 0:17:07.444 --> 0:17:10.564 full reporting, knowing everything that was going on, but let's 0:17:10.644 --> 0:17:15.444 keep that information from foreign governments. Now that's out right, 0:17:15.844 --> 0:17:19.204 So that's been rescinded. So what's and that's one of 0:17:19.204 --> 0:17:21.524 the main issues that we're seeing with deep seek. So 0:17:21.844 --> 0:17:23.404 so what do we think about that? What do we 0:17:23.404 --> 0:17:27.404 think about the government approach? Is it misguided or is 0:17:27.444 --> 0:17:29.684 it on the right track? And if we want to 0:17:29.844 --> 0:17:33.844 I think all of us, no matter if you're AAAI ORNAI, 0:17:34.204 --> 0:17:36.564 I think everyone wants to minimize pe doom, like I 0:17:36.564 --> 0:17:39.724 would hope that no one wants the world to be destroyed. 0:17:40.804 --> 0:17:44.244 So I don't think anybody thought that like this Biden 0:17:44.324 --> 0:17:48.684 executive order is going to stop pe doom by itself. 0:17:48.804 --> 0:17:51.764 The California law that was vetoed by Gavin Newsom probably 0:17:52.124 --> 0:17:54.324 was considered a bigger deal. It was a state law, 0:17:54.364 --> 0:17:56.364 but they're all based. If they want to operate in California, 0:17:56.364 --> 0:17:58.484 then they were subject to it, right, Like that might 0:17:58.524 --> 0:18:00.684 have been a bigger deal. But like the general pattern 0:18:00.724 --> 0:18:03.004 here with the California law failing, with this thing being rescinded, 0:18:05.044 --> 0:18:10.724 with Sam Altman being lesson less shall we say, concerned 0:18:11.564 --> 0:18:14.284 about safety, right, we're going to get this AI raise 0:18:14.404 --> 0:18:18.444 and this idea that you could stop it by having 0:18:19.604 --> 0:18:21.884 only you know, three operators until you reach some point 0:18:21.884 --> 0:18:26.964 where safety was achieved is not going to happen clearly, right, 0:18:28.244 --> 0:18:30.924 you know, Like, if you read any other technology, you 0:18:30.924 --> 0:18:33.084 would say, because one concern I have about AI or 0:18:33.204 --> 0:18:36.564 about this at the newsletter this week. You know, when 0:18:36.564 --> 0:18:39.004 you have these very big, powerful companies that have this 0:18:39.164 --> 0:18:42.604 lead in computing power and engineering talent, right, usually that's 0:18:42.604 --> 0:18:44.244 not how it works, Right, you have the next big thing, 0:18:44.284 --> 0:18:46.764 and the next big thing is created by new companies 0:18:46.804 --> 0:18:50.364 because the old companies are nostodgy and bloated and it's 0:18:50.364 --> 0:18:52.844 not their mission in the first place, and they're not cool, 0:18:52.884 --> 0:18:55.764 so don't attract young talent. Right, So to some accept 0:18:55.764 --> 0:18:59.204 the fact that, like, you know, it can be disrupted 0:18:59.284 --> 0:19:04.284 might lesson the worry about hegemonic concerns over AI and 0:19:04.324 --> 0:19:07.044 the kind of paternalistic slash people get very rich off 0:19:07.044 --> 0:19:09.684 it concerns about AI. But like, but yeah, I mean, look, 0:19:10.324 --> 0:19:15.004 I think we're a long way from achieving artificial superintelligence, 0:19:15.044 --> 0:19:17.844 which is the super human capabilities, right, But there are 0:19:17.884 --> 0:19:21.084 ways that AIS can be dangerous far short of this. 0:19:21.364 --> 0:19:21.564 Right. 0:19:21.684 --> 0:19:25.684 Take ching you how to like mix chemical compounds or 0:19:25.804 --> 0:19:28.684 build a pipe bomb or things like that, right, or 0:19:28.844 --> 0:19:31.724 can aid in a bit like sue with cidal thoughts 0:19:31.764 --> 0:19:33.524 and like in those use cases are going to be. 0:19:33.524 --> 0:19:35.404 Like they're already happening. 0:19:35.604 --> 0:19:39.404 They're already happening. We know that the guy who blew 0:19:39.484 --> 0:19:41.364 up the cyber truck in front of the Trump Hotel 0:19:42.444 --> 0:19:45.364 used chat GPT to figure out how to do it, 0:19:45.604 --> 0:19:47.924 which was actually probably one of the reasons it wasn't 0:19:47.964 --> 0:19:51.724 more destructive because the instructions were not very good. So 0:19:51.764 --> 0:19:53.844 I think I think we should be we should be 0:19:53.844 --> 0:19:58.404 grateful for the limitations of AI models for now. But yeah, no, 0:19:58.524 --> 0:20:01.684 there are I think I think that they're to me. 0:20:02.684 --> 0:20:04.884 That's actually one of the one of the more interesting 0:20:04.924 --> 0:20:07.404 points is that you know, we're worried about p doo, 0:20:07.604 --> 0:20:10.524 about you know, super intelligence, all these things, but I 0:20:10.564 --> 0:20:12.964 think that in the shorter term and potentially in the 0:20:13.004 --> 0:20:15.364 longer term, we need to be more worried about stupidity, 0:20:15.844 --> 0:20:20.044 right about the fact that uh that, like, they aren't 0:20:20.084 --> 0:20:25.524 super intelligent, right, and and people can misuse them, and 0:20:25.564 --> 0:20:29.564 they can give flawed outputs. And as they become more 0:20:29.644 --> 0:20:34.604 and more dominant mainstream used in searches, you know, used 0:20:34.644 --> 0:20:38.044 in used in day to day stuff, but also used 0:20:38.084 --> 0:20:41.644 higher up for people who you know, want something to 0:20:41.644 --> 0:20:44.764 summarize research for them, et cetera, et cetera, that the 0:20:45.364 --> 0:20:48.804 problem is going to be kind of much more mundane, 0:20:49.044 --> 0:20:51.124 right that it gives you that, it gives you bad information, 0:20:51.164 --> 0:20:55.844 it gives you bad instructions, It allides over something. It 0:20:55.844 --> 0:20:59.484 doesn't It doesn't quite synthesize something correctly. I think that 0:20:59.484 --> 0:21:04.004 that is actually the more pressing problem and is not 0:21:04.164 --> 0:21:06.444 P doom boom We're going to blow up, but is 0:21:06.484 --> 0:21:10.644 a kind of a smaller P dooms. It's already lowercase, 0:21:10.684 --> 0:21:15.004 but like subscript fe doom on a day to day basis, 0:21:16.004 --> 0:21:17.244 depending on who uses. 0:21:17.044 --> 0:21:19.764 It n for what. I've kind of flipped on this 0:21:19.804 --> 0:21:22.564 a little bit where I I think the hallucinations are 0:21:22.604 --> 0:21:25.204 an overrated problem. I think what the models are doing 0:21:25.284 --> 0:21:30.764 is is very impressive. They have fewer hallucinations than before. 0:21:32.324 --> 0:21:34.764 And you know, I use them a lot just for 0:21:34.884 --> 0:21:37.964 like research and problem solving a little bit of programming 0:21:38.004 --> 0:21:40.364 and things like that. And I think, you know, the 0:21:40.404 --> 0:21:42.124 one I've used the most is O one, which is 0:21:42.124 --> 0:21:47.724 the latest public build of chatgypt or open AI. It 0:21:47.804 --> 0:21:51.164 just can do higher level shit pretty well. You know, 0:21:51.964 --> 0:21:55.684 it also can catch itself in midstream. They put some 0:21:55.764 --> 0:21:59.444 routine in where it like well, typically a large language model. 0:21:59.684 --> 0:22:01.404 The reason why it seems like it's printing one word 0:22:01.404 --> 0:22:03.244 at a time. Is because like that's actually kind of 0:22:03.324 --> 0:22:08.644 like how it works, right. It literally it literally kind 0:22:08.644 --> 0:22:10.844 of goes to quench, It compresses its whole tex string, 0:22:10.844 --> 0:22:13.484 puts it through transformer. Right. But then you know then 0:22:13.564 --> 0:22:15.844 kind of it is recurs on how it puts it out. 0:22:16.004 --> 0:22:18.004 It doesn't think of it all at once, right. But 0:22:18.084 --> 0:22:20.844 now they've trained it to actually go back and check 0:22:20.884 --> 0:22:26.604 its output to check for hallucinations, and it catches them 0:22:26.884 --> 0:22:29.924 a lot of the time, but not a lot of 0:22:29.964 --> 0:22:30.284 the time. 0:22:30.364 --> 0:22:30.524 Right. 0:22:30.564 --> 0:22:34.084 So I actually just did some test runs prior to 0:22:34.364 --> 0:22:36.684 taping this so that I could see what was going 0:22:36.724 --> 0:22:40.084 on right now. And when it's in my area of expertise, 0:22:40.284 --> 0:22:44.324 I catch a lot of inaccuracies, not just hallucinations, but 0:22:44.404 --> 0:22:48.124 things that are almost right but kind of misunderstood the point, right, 0:22:48.164 --> 0:22:51.644 which which is actually which could be even more problematic. 0:22:51.724 --> 0:22:54.044 So I had to do some test runs in psychology 0:22:54.084 --> 0:22:55.124 where first of all. 0:22:55.004 --> 0:22:55.804 It did hallucinate. 0:22:55.844 --> 0:22:59.124 It still is making up studies, making up data that 0:22:59.164 --> 0:23:01.284 does not actually exist. When I try to, I'm like, oh, 0:23:01.284 --> 0:23:03.484 this is really cool. I want to look it up. No, 0:23:03.604 --> 0:23:07.284 it doesn't exist. But it also miss like it actually 0:23:07.284 --> 0:23:10.564 misinterprets findings and doesn't all always get it right. This 0:23:10.724 --> 0:23:12.884 is my expertise, right, I have a PhD at it, 0:23:13.004 --> 0:23:15.324 so I can figure this out and be like, you 0:23:15.324 --> 0:23:16.244 know what this is. 0:23:16.404 --> 0:23:17.964 It's pretty good, but not. 0:23:18.084 --> 0:23:21.244 Actually like you don't want to rely on this and 0:23:21.284 --> 0:23:25.404 this is just outright wrong. But because it's overall it 0:23:25.444 --> 0:23:27.884 seems pretty good, I don't catch it when it's not 0:23:27.964 --> 0:23:30.084 my area of expertise, and I just assume that it's 0:23:30.124 --> 0:23:32.684 pretty good and that pretty is doing a lot of 0:23:32.684 --> 0:23:33.364 heavy lifting. 0:23:33.884 --> 0:23:35.924 What about if you read a New York Times or 0:23:35.964 --> 0:23:40.284 Washington Post article on poker or sports betting, right or 0:23:40.404 --> 0:23:42.164 some feel bad, but. 0:23:42.124 --> 0:23:45.124 It doesn't actually hallucinate. It's not going to really it's 0:23:45.124 --> 0:23:47.404 not going to tell me about findings that don't exist 0:23:47.524 --> 0:23:48.284 to prove its point. 0:23:48.684 --> 0:23:50.324 It can tell a lot of white lies though, and 0:23:50.404 --> 0:23:52.484 misrepresent and be fundamentally dishonest. 0:23:53.044 --> 0:23:56.084 Right, Well, sure you have. You have a problem of 0:23:56.404 --> 0:23:59.644 reporting bias always, But there's a problem when you think 0:23:59.684 --> 0:24:02.684 something is factual information and it's not right. When I'm 0:24:02.684 --> 0:24:04.444 reading an op ed, I know it's an op ed 0:24:04.484 --> 0:24:07.404 when I'm reading an article, but when i'm when I 0:24:07.444 --> 0:24:09.164 want this for facts. 0:24:09.444 --> 0:24:11.524 But I want I want to know, not the use case. 0:24:11.564 --> 0:24:13.484 Though you can't put everything in the microwave oven. 0:24:13.764 --> 0:24:16.084 Of course I want to put one of the things 0:24:16.724 --> 0:24:18.844 if I want to learn about a field. So let 0:24:18.884 --> 0:24:21.044 me give you another example. I didn't just use psychology. 0:24:21.364 --> 0:24:25.564 I asked about I wanted to know about a specific 0:24:25.604 --> 0:24:27.164 type of company. I'm not going to give you kind 0:24:27.164 --> 0:24:30.964 of the search that was making a specific type of 0:24:30.964 --> 0:24:33.964 crypto investment. So I asked, like, what's a list of 0:24:33.964 --> 0:24:35.604 companies that's done it? And gave me a list. I 0:24:35.604 --> 0:24:38.084 was like, great, now will you tell me what the 0:24:38.084 --> 0:24:40.924 specific investments are? And it's like, oh, sorry, we don't 0:24:40.924 --> 0:24:42.884 actually know, Like we don't know that these companies have 0:24:42.924 --> 0:24:45.884 made any investments. We just gave you a list, And 0:24:46.284 --> 0:24:47.244 I was like, okay. 0:24:47.124 --> 0:24:52.484 You're using it wrong. You there are a lot of 0:24:52.484 --> 0:25:00.364 situations in life where precision isn't that important, right, you know? 0:25:00.564 --> 0:25:04.284 I mean, for example, I was as you made no listeners. 0:25:04.324 --> 0:25:06.604 I was in Korea and Japan recently, and like something 0:25:06.604 --> 0:25:09.364 I've been lazy about. I never taught myself how to 0:25:09.564 --> 0:25:14.164 distinguished Japanese, Chinese and Korean characters, and so I like 0:25:14.244 --> 0:25:17.924 told to bet I'm on the flight to Tokyo. I'm like, hey, 0:25:17.924 --> 0:25:20.804 give me a little very quick summary and give me 0:25:20.804 --> 0:25:22.084 a pop quiz and like you learn it in like 0:25:22.084 --> 0:25:24.044 ten or fifteen minutes, right, And they're like, you know, 0:25:24.044 --> 0:25:27.124 I don't have to like distinguish those characters one hundred accuracy. 0:25:27.164 --> 0:25:30.444 But it's like basically pretty good, and like I can 0:25:30.484 --> 0:25:34.844 tailor exactly how how that resources geared toward me. 0:25:36.124 --> 0:25:38.204 I'm not saying this is useless. I'm just saying that 0:25:38.244 --> 0:25:41.004 it's right now. It is being used for the cases 0:25:41.044 --> 0:25:42.844 that I've told you about because it's at the top 0:25:42.884 --> 0:25:46.964 of your such when you do a Google search, like 0:25:47.644 --> 0:25:49.604 that's the first thing that comes up as the. 0:25:49.724 --> 0:25:54.524 I think it's very bad branding decision by by Google, right. 0:25:54.564 --> 0:25:56.644 I do think it's a bad because like and Google's 0:25:56.764 --> 0:26:00.044 I'm sorry, it's not as good as opening Eye show anthropic. 0:26:00.084 --> 0:26:03.964 It's not it's not sorry Google, and like it undermines 0:26:04.044 --> 0:26:06.204 Google's kind of lead in search. And I think Google 0:26:06.204 --> 0:26:08.244 has handled this a lot of things in this space 0:26:08.364 --> 0:26:10.764 very badly. Although I mean, you know, it was Google 0:26:10.804 --> 0:26:13.164 engineers who came up with transformer paper and they and 0:26:13.204 --> 0:26:17.244 they and they, like you know, still hire lots of 0:26:17.244 --> 0:26:18.724 great pale up but they've kind of become like this 0:26:18.804 --> 0:26:23.124 feeder system to like the hip or cooler AI companies, 0:26:23.444 --> 0:26:26.124 I think, But like anyway I think people are I 0:26:26.164 --> 0:26:30.044 think people are way too hipster about this, Like this 0:26:30.084 --> 0:26:33.444 is the most quickly adopted technology in the history of 0:26:33.444 --> 0:26:34.844 the world by some mess. 0:26:34.684 --> 0:26:37.644 Absolutely, and I'm not like I actually like, I think 0:26:37.684 --> 0:26:40.284 that it has a lot of potential. I want it 0:26:40.324 --> 0:26:43.444 to do better, right, Like I like I want them 0:26:43.444 --> 0:26:44.284 to fix this shit. 0:26:44.764 --> 0:26:44.964 Right. 0:26:45.044 --> 0:26:47.404 Look, we are poker players, Maria. We should be used 0:26:47.444 --> 0:26:51.724 to being able to accept information as shifting your prior 0:26:51.804 --> 0:26:54.684 or shifting your view, but like not being definitive, right, 0:26:54.724 --> 0:26:57.124 And that's a journalist. We're both journalists too, Like you know, 0:26:57.244 --> 0:26:59.004 when a source tells you something, you've vet. 0:26:59.004 --> 0:27:03.004 It absolutely, But it actually adds more work for me 0:27:03.444 --> 0:27:05.444 because I have to go through it and try to 0:27:05.484 --> 0:27:07.644 figure out what can I rely on, what can't I 0:27:07.724 --> 0:27:11.844 rely on. Now, one listener who who has shared his 0:27:11.924 --> 0:27:15.524 experience on UH on social media as well, and I 0:27:15.564 --> 0:27:18.884 know you wrote you kind of referenced it in your 0:27:18.884 --> 0:27:22.524 newsletter this week, Kevin Ruse had emailed me about poker 0:27:22.564 --> 0:27:26.364 training and said that he was using chat GPT and 0:27:26.444 --> 0:27:28.764 AIS to help him with poker. And I said, don't 0:27:28.764 --> 0:27:31.324 do that because it's going to tell you the wrong thing. 0:27:32.284 --> 0:27:36.364 And he did it. He still did it and he said, oh, 0:27:36.364 --> 0:27:38.484 I want to tournament. It was it was good, it 0:27:38.564 --> 0:27:41.524 was helpful. So I actually had chat GPT do some 0:27:41.604 --> 0:27:42.564 poker training for me. 0:27:42.884 --> 0:27:43.964 It's not good. 0:27:44.404 --> 0:27:46.844 It gives you incorrect advice if you don't know, if 0:27:46.844 --> 0:27:49.164 you're a novice, and if you're using this, you might 0:27:49.204 --> 0:27:52.284 get lucky, right, and it were all works out, but 0:27:52.404 --> 0:27:55.644 let's go to poker, like, try to use chat GPT 0:27:55.684 --> 0:27:58.404 to teach you poker strategy. It is not going to 0:27:58.404 --> 0:28:01.604 teach you strategy. Did you did you and the better 0:28:01.644 --> 0:28:01.924 you are? 0:28:03.804 --> 0:28:04.004 Yeah? 0:28:04.444 --> 0:28:06.564 Yeah, I just I did the same and I thought 0:28:06.564 --> 0:28:09.724 it was pretty good. It was it misses things that 0:28:09.804 --> 0:28:12.124 like so I actually did this last night. I mean 0:28:12.164 --> 0:28:14.444 it gets you know, I so what do you? What 0:28:14.444 --> 0:28:14.844 do you actually? 0:28:15.244 --> 0:28:19.284 But here's the thing, though, are you yellous? You're you're able. 0:28:19.044 --> 0:28:22.124 To distinguish what it's missing and what is getting pretty well. 0:28:22.364 --> 0:28:25.724 If you're using this as the tool to train yourself 0:28:26.244 --> 0:28:30.684 and you don't have any background knowledge, that is the problem, right, 0:28:30.724 --> 0:28:33.244 you need it to be you need it to teach 0:28:33.244 --> 0:28:36.644 you correctly. It's much more difficult as someone who started 0:28:36.684 --> 0:28:40.804 poker from zero as an adult. Right, let me tell you, Like, 0:28:40.924 --> 0:28:42.804 one of the most important things I learned was that 0:28:42.844 --> 0:28:46.004 it's much easier to teach someone from zero because I 0:28:46.004 --> 0:28:47.204 didn't have any bad habits. 0:28:47.284 --> 0:28:47.444 Right. 0:28:47.484 --> 0:28:51.004 If I had instead learned from chat GPT and those 0:28:51.084 --> 0:28:53.844 were kind of the habits and the thought processes that 0:28:53.884 --> 0:28:56.564 I acquired, and some of them were just wrong or 0:28:56.604 --> 0:28:59.484 didn't teach me how to think correctly through things, I'd 0:28:59.484 --> 0:29:01.084 be a really shitty poker player. 0:29:01.684 --> 0:29:03.604 Yeah, so let me give you some example of how 0:29:03.644 --> 0:29:05.444 I use chat GPT. 0:29:05.804 --> 0:29:05.964 Right. 0:29:07.244 --> 0:29:11.924 You know, one is kind of as a research assistant, 0:29:12.004 --> 0:29:14.324 but like once you already know something about a topic, right, 0:29:14.324 --> 0:29:15.964 Like I'm not getting a first brief, but I'm like 0:29:16.044 --> 0:29:18.804 quarrying it where I'm saying, Okay, I talked to this person. 0:29:19.924 --> 0:29:24.044 Here's a description of how an AI thing works, or 0:29:24.084 --> 0:29:26.644 a crypto thing works, or concept and finance works. Right, 0:29:27.004 --> 0:29:30.684 will you vet this for me? What critiques might you have? Right? 0:29:31.604 --> 0:29:33.524 It's maybe not quite as good as talking to like 0:29:33.564 --> 0:29:35.684 an expert, but I find like you often get a 0:29:35.724 --> 0:29:37.404 lot of value from that, and again it's not the 0:29:37.444 --> 0:29:38.924 last step in the process. 0:29:39.004 --> 0:29:39.124 Right. 0:29:39.124 --> 0:29:41.804 You can also use it for creative inspiration. Give me 0:29:41.924 --> 0:29:44.444 ten potential headlines from this. You can use to fill 0:29:44.484 --> 0:29:46.004 in missing words, because it thinks in terms of a 0:29:46.004 --> 0:29:48.724 big matrix, right, so like what's an analogy that I 0:29:48.764 --> 0:29:51.844 can think of? What's this word or concept that I'm missing? 0:29:51.884 --> 0:29:55.044 Invent the name for this thing or that thing. You 0:29:55.044 --> 0:29:57.364 can use it to kind of squeeze quantitative data out 0:29:57.364 --> 0:29:59.764 of qualitative information, like I asked it, for example, to 0:30:00.564 --> 0:30:02.604 an identify straft to this again, right, but to like 0:30:03.524 --> 0:30:07.564 vet my estimate of how liberal or conservative different eras 0:30:07.564 --> 0:30:10.204 in American history we're on a negative to to positive 0:30:10.244 --> 0:30:13.324 ten scale. It can make ranking lists of different kinds. 0:30:13.324 --> 0:30:15.324 I mean, it's just like there are so many use 0:30:15.324 --> 0:30:19.964 cases for it, and like people just want us to 0:30:19.964 --> 0:30:23.084 like cheat on papers or use it as a substitute 0:30:23.084 --> 0:30:24.804 for like Wikipedia or something, which you're not the best 0:30:24.924 --> 0:30:26.964 use cases for it. And if you ask chet ChiPT, 0:30:27.124 --> 0:30:28.724 it will tell you that those are not the best 0:30:29.164 --> 0:30:30.044 use cases for it. 0:30:30.164 --> 0:30:30.324 Right. 0:30:30.404 --> 0:30:32.964 The queriable nature of it and the fact that it 0:30:33.044 --> 0:30:35.564 reorganizes this information in a way that for many purposes 0:30:35.564 --> 0:30:38.684 but not all purposes, is much more approachable accessible, is 0:30:38.804 --> 0:30:41.204 it's I don't know, it's I think it's a miraculous technology. 0:30:41.444 --> 0:30:44.684 I mean, you know you had woken up. If I 0:30:44.724 --> 0:30:50.004 had fallen to a coma years in twenty fifteen and 0:30:50.284 --> 0:30:55.604 woken up, I missed the whole first Trump administration, I'm like, oh, 0:30:55.604 --> 0:31:00.284 Trump's prison, Oh he was already President's surprise then, like 0:31:00.644 --> 0:31:02.844 you would be fucking blown away by this shit, right, 0:31:02.844 --> 0:31:04.964 and be like, oh my fucking god. Right, just like 0:31:05.004 --> 0:31:07.764 passing the Turing test. I mean, there are definitions to 0:31:07.764 --> 0:31:08.924 meet it's but whether it's a good test, whether it 0:31:09.004 --> 0:31:13.324 actually is, But like it basically is like human esque 0:31:13.444 --> 0:31:18.564 intelligence in some ways, inferior in some cases superior over 0:31:18.644 --> 0:31:20.844 a large domain of fields. Just the way it can 0:31:21.004 --> 0:31:26.164 like parse this very open ended, fuzzy logic of techt strings. 0:31:26.204 --> 0:31:27.804 I mean, I just think it's kind of an amazing 0:31:28.644 --> 0:31:32.404 It's amazically a robust in some ways, right, sure that 0:31:32.444 --> 0:31:35.204 you can misspell things you get, I mean, it's amazingly 0:31:35.284 --> 0:31:37.684 robust in a way that like it's hard to think 0:31:37.684 --> 0:31:43.564 of other technologies that compare to it exactly, and you know, 0:31:43.804 --> 0:31:49.004 and solved using very simple underlying math, right, I mean, 0:31:49.044 --> 0:31:51.124 the code for deep seek is something like a few 0:31:51.204 --> 0:31:54.444 hundred lines of code long. My fucking election model is 0:31:54.524 --> 0:31:58.084 longer than that. Right, It's like it's like it's kind 0:31:58.084 --> 0:31:58.724 of a miracle. 0:31:59.084 --> 0:32:04.924 Absolutely absolutely I agree with all of that. Where but 0:32:05.244 --> 0:32:08.324 now I think to push it a little bit further. Obviously, 0:32:08.364 --> 0:32:12.564 this miracle, as we know, it's coming at a big cost, right, 0:32:12.924 --> 0:32:17.444 environmental cost. It also P doom. Right, we've talked about 0:32:17.524 --> 0:32:21.164 that potential risk is the miracle of you know, giving 0:32:21.164 --> 0:32:25.164 you a good analogy worth it at this cost. That's 0:32:25.204 --> 0:32:26.164 I think those are kind. 0:32:26.084 --> 0:32:28.524 Of don't give me this environmental. 0:32:28.324 --> 0:32:31.284 I'm not talking I'm talking about P doom and environmental. 0:32:31.284 --> 0:32:34.164 It's not crap, Nate. We talked about this before, like 0:32:34.604 --> 0:32:35.444 a lot of energy. 0:32:35.524 --> 0:32:37.324 Okay, look at this. This is something that's supposed to scare me. 0:32:37.524 --> 0:32:41.564 Four impress dot org. The energy consumption for training chat 0:32:41.604 --> 0:32:45.244 EPT leading model is even more staggering, equated to that 0:32:45.284 --> 0:32:47.724 of an American hassole for more than seven hundred years. 0:32:47.764 --> 0:32:51.324 So basically, to train this leading model only took seven 0:32:51.404 --> 0:32:56.244 hundred households worth like one subdivision of some fucking neighborhood 0:32:56.284 --> 0:33:00.044 in Tulsa. Right, it's not very much like don't you 0:33:00.084 --> 0:33:02.444 know you undermind the argument for ped doom. We all 0:33:02.484 --> 0:33:05.084 die if like I mean this, you know, obviously if 0:33:05.764 --> 0:33:07.804 the models get hungrier and hungrier. But by the way, 0:33:07.804 --> 0:33:11.844 the deep seek thing should be good news for the environment. 0:33:11.964 --> 0:33:14.124 Well that's why I said, we don't know, because on 0:33:14.164 --> 0:33:17.644 the one hand, maybe depending on what the actual resources are. 0:33:17.844 --> 0:33:19.764 On the other hand, if it means that every single 0:33:19.764 --> 0:33:22.164 person is now running these smaller things, or not every 0:33:22.164 --> 0:33:25.364 single person, but if it makes it more likely that 0:33:25.484 --> 0:33:28.204 more of these are what's the net impact, right, If 0:33:28.244 --> 0:33:30.964 it's one tenth but you actually have one hundred times 0:33:31.204 --> 0:33:33.724 more people using it as a result, then it's a 0:33:33.764 --> 0:33:37.324 net obviously negative impact instead of net positive. These are 0:33:37.364 --> 0:33:39.764 all open questions, right, And like I said, I am 0:33:39.844 --> 0:33:42.524 not an AI skeptic. I think it's really cool. I 0:33:42.564 --> 0:33:45.044 think there are lots of really interesting things here. I 0:33:45.124 --> 0:33:47.044 just think that there are other you know, you can't 0:33:47.084 --> 0:33:49.364 also be like a raw raw cheer later none of 0:33:49.364 --> 0:33:52.004 this matters. Of course, it does matter. I think all 0:33:52.044 --> 0:33:52.844 of these things matter. 0:33:53.204 --> 0:33:56.884 I just think people are over indexing to like I 0:33:56.884 --> 0:33:59.724 don't know. I mean, have you taken wes Wimo Maria 0:33:59.724 --> 0:34:01.284 and by the way if you're not aware, this is 0:34:01.324 --> 0:34:06.604 a self driving car company which is available in San Francisco, Phoenix, 0:34:06.644 --> 0:34:08.044 and maybe one or two other places. 0:34:08.124 --> 0:34:08.284 La. 0:34:08.364 --> 0:34:10.204 I think have you taken a Weimo in any. 0:34:10.044 --> 0:34:12.324 Of those places, Maria, I have not name. 0:34:12.764 --> 0:34:15.244 It's fucking blade wrinner. I'm telling you, it's a very 0:34:15.284 --> 0:34:17.724 good experience. It's a much smoother ride than i'd say 0:34:17.764 --> 0:34:21.364 ninety five percent of ubers. They have like space sage music, 0:34:21.404 --> 0:34:24.124 and you feel like you're in the fucking future. And 0:34:24.324 --> 0:34:29.324 I would almost guarantee you that driver less cars are 0:34:29.324 --> 0:34:32.204 going to be a very popular technology. 0:34:32.324 --> 0:34:33.164 I don't know. I don't know. 0:34:33.804 --> 0:34:36.364 I watched I watched that. I watched that episode of 0:34:36.404 --> 0:34:39.044 Silicon Valley where he's in the driver less car and 0:34:39.164 --> 0:34:42.044 en step on a on a boat somewhere in the 0:34:42.084 --> 0:34:47.404 middle of the ocean. So I'm a little obviously TV 0:34:47.484 --> 0:34:48.364 show comedy. 0:34:48.444 --> 0:34:51.364 But but you know, you never you never know how 0:34:51.364 --> 0:34:53.164 the experience will will end up. 0:34:53.284 --> 0:34:55.404 But I'm going to San Francisco next week, Nate, so 0:34:55.604 --> 0:34:57.644 you know, maybe I'll take my first Weimo. 0:34:57.644 --> 0:35:00.164 Take a Wimo. It's it's it's like a ninety fifth 0:35:00.244 --> 0:35:02.764 percentile Uber driver. 0:35:03.324 --> 0:35:06.324 All right, Well, well, on that positive note, Shall we 0:35:06.444 --> 0:35:08.484 talk a little bit more poker and switch to a 0:35:08.524 --> 0:35:16.844 listener question? Okay, fine, we'll be back right after this. 0:35:27.604 --> 0:35:32.684 All right, So we had a poker related listener question 0:35:33.004 --> 0:35:38.364 that I think, Nate, you are probably more equipped to 0:35:38.404 --> 0:35:41.404 answer in the sense that you play cash home games 0:35:41.444 --> 0:35:43.964 and I don't. By the way, I'm really sorry if 0:35:44.004 --> 0:35:47.964 you can hear some knock knock noises in the background. 0:35:48.724 --> 0:35:52.804 Apparently the apartment above mine has just started construction. It 0:35:52.884 --> 0:35:56.444 actually just happened as we started taping this podcast. This 0:35:56.604 --> 0:35:59.404 is the first hammering I have heard. But of course 0:35:59.484 --> 0:36:02.724 you get to experience it alongside me, because I love 0:36:02.724 --> 0:36:04.804 our listeners and I want to share all of my 0:36:04.924 --> 0:36:09.204 experiences with them. So Nate, here's the listener question. I 0:36:09.284 --> 0:36:12.204 have a neighborhood poker night with my friends. Everyone plays 0:36:12.204 --> 0:36:14.924 really loose and passive, lots of calling, not much raising. 0:36:15.364 --> 0:36:18.004 How do I win against real amateurs like that? What 0:36:18.084 --> 0:36:20.404 are the most common and easy to detect tells by 0:36:20.444 --> 0:36:23.644 amateurs like this? So part of this I can answer too, right, 0:36:23.644 --> 0:36:26.164 because this happens in tournaments as well. But let's start 0:36:26.164 --> 0:36:29.844 with what you think. Since you play in home games, 0:36:29.924 --> 0:36:32.964 you know this is something that you find fun and 0:36:33.564 --> 0:36:35.764 not an experience that I often have. 0:36:37.564 --> 0:36:39.884 So it depends this comes from listener. 0:36:40.044 --> 0:36:40.204 Hugh. 0:36:40.524 --> 0:36:46.924 H u g h. I guess that's the way you spill, Hugh. 0:36:48.684 --> 0:36:49.764 You does sound British. 0:36:49.844 --> 0:36:52.884 I'm sorry, it's just a name that I automatically associate 0:36:52.924 --> 0:36:53.924 with you with being English. 0:36:54.444 --> 0:36:57.844 We have Okay, So what are basics for loose home? 0:36:57.884 --> 0:37:00.364 I mean depends on if you're talking about like really 0:37:00.404 --> 0:37:02.804 bad players. He seems to be. 0:37:03.044 --> 0:37:05.364 He seems to be, so. 0:37:06.164 --> 0:37:08.764 You know, the basics are you actually don't want to 0:37:08.764 --> 0:37:10.884 play like everyone else beating, you know, don't be so 0:37:11.084 --> 0:37:16.724 loose and passive. Right, you need particularly out of position 0:37:17.044 --> 0:37:21.804 hands that can make the nuts right, so suited hands 0:37:21.844 --> 0:37:26.804 particularly you know asex suited, Broadway suited hands ten nine 0:37:26.844 --> 0:37:29.444 suited kind of an above right. So number one, hand 0:37:29.484 --> 0:37:34.884 selection becomes more more oriented toward strong hands that can 0:37:34.884 --> 0:37:39.404 make big you know, straights and flushes and better. That's 0:37:39.484 --> 0:37:42.404 part one, right, Number two, you're gonna want to like 0:37:42.924 --> 0:37:46.924 increase your bet sizing maybe quite a bit. Theory says 0:37:46.924 --> 0:37:48.524 in a cash game that you're supposed to raise to 0:37:48.564 --> 0:37:50.764 maybe two and a half x a big blind here 0:37:50.804 --> 0:37:53.564 you can go four or five if they're already limpers. 0:37:53.684 --> 0:37:56.444 You can raise even more than that. Right, There are 0:37:56.444 --> 0:37:58.284 some games where the standard open might be to like 0:37:58.644 --> 0:38:00.204 ten x or things like that. But like, let me 0:38:00.244 --> 0:38:02.644 maybe let me even back up a little further, right, 0:38:02.764 --> 0:38:05.684 I actually have dealt poker games to total rank amateurs 0:38:05.764 --> 0:38:07.484 or like literally five to ten people. I have never 0:38:07.524 --> 0:38:08.764 played booker before. 0:38:08.844 --> 0:38:09.044 Right. 0:38:11.404 --> 0:38:14.404 The two things that they most routinely get wrong are 0:38:14.644 --> 0:38:18.724 number one, they call too much, meaning they call and 0:38:18.804 --> 0:38:21.124 play it like a slot machine instead of folding or 0:38:21.164 --> 0:38:25.244 raising more. And number two, they don't understand bet sizing. 0:38:25.764 --> 0:38:27.604 What is the size of your bet relative to the 0:38:27.684 --> 0:38:31.844 size of the pot. Right, If you don't know anything 0:38:31.844 --> 0:38:35.884 about poker, no, nothing at all, then just bet half 0:38:35.924 --> 0:38:38.484 the size of the pot. Keep track of what's in 0:38:38.524 --> 0:38:41.084 the pot and bet half that size. But in general, 0:38:41.404 --> 0:38:42.444 don't do so much calling. 0:38:42.484 --> 0:38:42.604 Right. 0:38:42.644 --> 0:38:44.204 If you have a good hand or a good bluff, 0:38:44.444 --> 0:38:45.684 or even if you just kind of think other people 0:38:45.724 --> 0:38:49.964 are scared, then then do some raising if you think 0:38:50.004 --> 0:38:52.084 there'll be then most amateur players are not going to 0:38:52.164 --> 0:38:53.764 go nuts. I mean it's a little complicate because like 0:38:54.204 --> 0:38:56.404 they might not understand hands strikes, so you might want 0:38:56.404 --> 0:38:58.324 to weig absolute strength a little bit more, but like, 0:38:59.324 --> 0:39:00.884 but don't be afraid to get the money. And when 0:39:00.924 --> 0:39:02.644 you have a good hand, and when other people have 0:39:02.684 --> 0:39:06.404 a good representing a good hand, then it gets a 0:39:06.444 --> 0:39:10.564 little bit complicated. But then you know, you don't have 0:39:10.604 --> 0:39:13.764 to do a whole excess amount of bluff catching. I mean, 0:39:13.764 --> 0:39:17.524 those are the basic And then I'd say, like, in general, 0:39:17.564 --> 0:39:20.404 in these lose cash games, people are very sticky. Now 0:39:20.404 --> 0:39:22.924 we're talking about a slightly higher caliber of players people 0:39:22.964 --> 0:39:25.924 have had played before. Right, people are mostly very sticky 0:39:25.964 --> 0:39:29.644 pre flop and on the flop, and then they will 0:39:29.724 --> 0:39:32.484 start to fold cash can players do like to fold 0:39:32.524 --> 0:39:35.404 on turns and rivers sometimes, Right, So that means that, 0:39:35.484 --> 0:39:37.964 like you know that can affect your whole strategy for 0:39:38.004 --> 0:39:39.484 the whole hand is that you tend not to have 0:39:39.524 --> 0:39:41.644 a lot of full equity preflop in on flops, and 0:39:41.684 --> 0:39:43.924 then it requires multiple barrels sometimes. 0:39:44.364 --> 0:39:45.124 Yeah. 0:39:45.204 --> 0:39:48.604 I think that as someone who is a tournament player, 0:39:49.604 --> 0:39:52.964 there is some advice that I think applies all around, 0:39:53.684 --> 0:39:56.964 which basically goes hand in hand with what you said Nate. 0:39:57.804 --> 0:40:00.484 Number one, You don't want to follow the tendencies of 0:40:00.564 --> 0:40:03.084 the people who are making mistakes, right, So If people 0:40:03.124 --> 0:40:05.564 are too loose, you actually want to tighten up. If 0:40:05.604 --> 0:40:09.564 people are passive, you want to become more aggressive. I 0:40:09.604 --> 0:40:11.924 think that's important. But you should also realize if they're 0:40:11.964 --> 0:40:14.724 going to be sticky, then you should just bet huge, right, 0:40:14.844 --> 0:40:17.404 Like if you are going to if you'd normally bet 0:40:17.404 --> 0:40:20.284 half pot, just bet pot they're going to call anyway, right, 0:40:20.724 --> 0:40:23.964 build massive pots with your good hands. This will also 0:40:24.084 --> 0:40:28.484 enable you to bluff, right, because they will eventually fold. Now, 0:40:28.684 --> 0:40:30.684 something you said, I think this is actually true, not 0:40:30.804 --> 0:40:35.044 just of cash games, but in general in tournaments as well. 0:40:35.604 --> 0:40:41.204 People do tend to overcall flop and overfold turn. So 0:40:41.364 --> 0:40:44.444 that's a great you know, I think building an over 0:40:44.484 --> 0:40:47.644 betting strategy into your game in a game like that 0:40:47.764 --> 0:40:50.884 is really good, right, And sometimes they'll get very sticky 0:40:51.084 --> 0:40:52.924 like I've had. Because the second part of this question 0:40:53.044 --> 0:40:55.404 was tells, which I think is just bad. Do not 0:40:55.564 --> 0:40:58.324 use tells, even though in games like that people probably 0:40:58.364 --> 0:41:00.764 do have tells, especially if you're going to play with 0:41:00.804 --> 0:41:02.564 them over and over. It might be a little different, 0:41:02.804 --> 0:41:04.644 but I just don't think that's great to rely on. 0:41:05.004 --> 0:41:07.884 But when I've relied on tells, I've actually made really 0:41:07.924 --> 0:41:11.724 big mistakes. Because I've had I've had a situations where 0:41:11.724 --> 0:41:13.764 I'm like, oh, this person really likes their hand, they 0:41:13.844 --> 0:41:16.924 must be really strong, and I end up folding. And 0:41:17.044 --> 0:41:19.484 they had like ace deucee off suit, but there was 0:41:19.484 --> 0:41:21.044 an ace on the board, and they thought that it 0:41:21.084 --> 0:41:24.484 was just like the nuts, right, because they completely overvalued 0:41:24.524 --> 0:41:26.924 the fact that they had an ace, and so they 0:41:26.924 --> 0:41:28.564 were playing it like they had the nuts and they 0:41:28.604 --> 0:41:30.884 thought they had the nuts, but they really didn't. So 0:41:30.924 --> 0:41:34.124 if people are bad, don't use tells because the strength 0:41:34.764 --> 0:41:37.524 their perceived strength of their hand may not actually be 0:41:37.604 --> 0:41:39.564 the actual strength of their hands. 0:41:39.604 --> 0:41:44.204 I'm more till it's funny because we had like opposite personnelity. 0:41:44.204 --> 0:41:46.404 You're like way more sound theoretically, and I'm kind of 0:41:46.444 --> 0:41:51.364 like psychoanalyzing people a little bit more. And no, look, 0:41:51.404 --> 0:41:55.684 I think the premise are like two categories of tells 0:41:55.724 --> 0:42:00.724 from very inexperienced players that I don't think I're always 0:42:00.724 --> 0:42:03.484 hard to distinguish, but require additional context. Right, what is 0:42:03.524 --> 0:42:08.204 a really bad actor tell right where they just are 0:42:08.244 --> 0:42:10.564 like they watched like poker movies, We're supposed to act 0:42:11.084 --> 0:42:12.604 if you're strong and starting every week and they just 0:42:12.644 --> 0:42:16.044 really over knew it. In like a comical way, right, yeah, 0:42:16.044 --> 0:42:19.524 and then I have seen that, but just sometimes people 0:42:19.564 --> 0:42:22.444 are like extremely There's also there's a. 0:42:22.324 --> 0:42:25.684 Lot of hollywooding actually, and I've seen this in uh 0:42:25.804 --> 0:42:28.524 from amateur where like if you have the nuts right, 0:42:28.564 --> 0:42:32.284 like say you flopped quads right or something like that, 0:42:32.564 --> 0:42:37.724 and then they'll just be like the size and then 0:42:37.764 --> 0:42:41.644 like I guess I call right like that if someone's 0:42:41.724 --> 0:42:43.884 doing that, like holy shit, you're beat. 0:42:44.604 --> 0:42:47.964 Like just there are there are certain situations like. 0:42:47.924 --> 0:42:51.444 That, but I think in general it's it's better to 0:42:51.884 --> 0:42:53.004 we haven't played. 0:42:52.764 --> 0:42:55.324 In Hughes game or Hughes game. 0:42:55.964 --> 0:42:59.764 So I think that just sticking to the advice that 0:42:59.764 --> 0:43:04.564 that we've given, which is, you know, don't be lose passive. Basically, 0:43:04.764 --> 0:43:06.324 you have to tighten up your ranges. You have to 0:43:06.324 --> 0:43:11.364 figure out what those ranges are. And you know, the 0:43:11.364 --> 0:43:14.284 the other thing is you know your bet sizing is 0:43:14.324 --> 0:43:16.004 going to change, because if people are going to be 0:43:16.044 --> 0:43:18.924 calling stations, great, exploit it. If people are going to 0:43:18.924 --> 0:43:22.124 call pre flop anyway, great, make your sizes bigger. Just 0:43:22.604 --> 0:43:25.404 build pots when you have very strong hands. 0:43:25.484 --> 0:43:27.204 I mean the other thing, you know, to close this 0:43:27.284 --> 0:43:30.884 cussing on tells I mean, people can also be very honest, right, 0:43:31.004 --> 0:43:35.764 Like they don't you know, in games where the stakes 0:43:35.804 --> 0:43:38.964 are low relative to people's like net worth, which depends 0:43:39.004 --> 0:43:42.444 on people's net worth, right, then they just don't necessarily 0:43:42.524 --> 0:43:48.004 take a lot of action to like conceal disappointment with 0:43:48.084 --> 0:43:52.404 a bad flop or things like that. You know a 0:43:52.444 --> 0:43:53.924 lot of times all I gotta catch my car like 0:43:53.964 --> 0:43:57.564 that actually often is more than not honest, more in 0:43:57.644 --> 0:44:00.004 cash games and in tournaments. I don't know why, right, 0:44:01.244 --> 0:44:04.644 I think tournaments people like are just like playing their 0:44:04.844 --> 0:44:09.604 A game a bit more Amit's secret of cash right more? 0:44:09.604 --> 0:44:12.604 Playing their A game more often in tournaments at games. 0:44:12.524 --> 0:44:15.004 Yeah, no, I actually I think I think there's something 0:44:15.004 --> 0:44:17.084 too that people do tend to be more honest in 0:44:17.124 --> 0:44:19.964 cash games. I had a hilarious situation at a higher 0:44:19.964 --> 0:44:23.724 stakes cash game where I had raised and I don't 0:44:23.724 --> 0:44:25.804 remember if it was small or big blind had defended 0:44:26.844 --> 0:44:32.244 and anyway, went, you know, check bet on the flop, 0:44:33.404 --> 0:44:38.604 then check and check on the turn and check and no, no, 0:44:38.644 --> 0:44:41.324 not check And I was like looking to see like 0:44:41.444 --> 0:44:43.444 I had nothing what I wanted to bet on the 0:44:43.524 --> 0:44:46.164 river and he just folded. He's like, you definitely have 0:44:46.284 --> 0:44:49.444 me beat, because like I've got nothing and I had nothing, right, Like, 0:44:49.484 --> 0:44:51.244 I don't think I definitely had a beat, and he 0:44:51.364 --> 0:44:53.284 just folded to me, right. I didn't even have to 0:44:53.284 --> 0:44:55.084 think about the sizing or whether I was going to 0:44:55.124 --> 0:44:57.444 bet or any of it. That would never happen in 0:44:57.484 --> 0:45:00.124 a tournament, but it happens in cash games all the time. 0:45:00.164 --> 0:45:01.284 And I still remember this hand. 0:45:01.284 --> 0:45:03.964 I don't play cash very often, so things like that 0:45:04.044 --> 0:45:06.364 stand out, but I've seen people do that and then 0:45:06.364 --> 0:45:07.964 they try to do it in tournaments. Actually, you can 0:45:08.004 --> 0:45:10.724 often spot a cash player a tournament because they will 0:45:10.764 --> 0:45:13.244 sometimes fold out of turn. They'll do things that are 0:45:13.284 --> 0:45:16.484 just like very honestly communicate that they have no more 0:45:16.524 --> 0:45:17.204 interest in. 0:45:17.084 --> 0:45:21.364 This hand and don't be a super knit right, like 0:45:21.444 --> 0:45:25.404 to show people over value, especially in cash games, the 0:45:25.484 --> 0:45:27.604 last thing they saw, right, So like if you have 0:45:27.724 --> 0:45:32.444 like an occasional hand where you get a little out 0:45:32.444 --> 0:45:36.004 of line, right, and again, I think you know usually 0:45:36.164 --> 0:45:38.804 worth picking your spots carefully, and there's some psychology to that. 0:45:38.884 --> 0:45:40.964 And then like if you show, oh, I three bit 0:45:41.084 --> 0:45:44.684 five four suited from the button, which might actually be 0:45:44.724 --> 0:45:48.604 a perfectly fine near gto three bed occasionally right, if 0:45:48.644 --> 0:45:52.164 you turn as straight with that an the other guy folds. 0:45:52.204 --> 0:45:53.764 You definitely want to show that hand right. You want 0:45:53.764 --> 0:45:57.604 to like you're maintaining your reputation because, like a seat 0:45:57.644 --> 0:45:59.484 in a good cash game is a valuable thing and 0:45:59.524 --> 0:46:02.644 people absolutely will notice if you're being a knit. Don't 0:46:02.644 --> 0:46:03.044 be a knit. 0:46:03.364 --> 0:46:03.564 Yep. 0:46:03.604 --> 0:46:06.524 Have fun, play toward the looser end of your GTO range. 0:46:06.844 --> 0:46:09.004 How your GT arrange may actually be pretty tight against 0:46:09.684 --> 0:46:12.964 phishing number fold yep, good luck you, good luck you. 0:46:19.764 --> 0:46:21.924 Let us know what you think of the show. Reach 0:46:22.004 --> 0:46:26.484 out to us at Risky Business at pushkin dot Fm. 0:46:26.644 --> 0:46:30.204 Risky Business is hosted by me Maria Kondikova and byby 0:46:30.484 --> 0:46:31.004 Nate Silver. 0:46:31.844 --> 0:46:35.364 The show is a co production of Pushkin Industries and iHeartMedia. 0:46:35.884 --> 0:46:39.444 This episode was produced by Isabel Carter. Our associate producer 0:46:39.524 --> 0:46:43.044 is Gabriel Hunter Chang. Our executive producer is Jacob Goldstein. 0:46:43.484 --> 0:46:45.604 If you like the show, please rate and review us 0:46:45.604 --> 0:46:47.684 so other people can find us too. And if you 0:46:47.724 --> 0:46:50.084 want to listen to an ad free version, sign up 0:46:50.084 --> 0:46:52.484 for Pushkin Plus. For We're six, ten and nine a 0:46:52.524 --> 0:46:55.004 month you get access to ad free listening. Thanks for 0:46:55.044 --> 0:46:55.444 tuning in.