WEBVTT 00:00:00.560 --> 00:00:02.799 Hey, what's up, Toby? We made it another week. 00:00:03.459 --> 00:00:05.339 Nice. Hey, Brian. How's it going? I hope we've 00:00:05.339 --> 00:00:07.320 got another topic. I think it's your week to 00:00:07.320 --> 00:00:11.880 come up with a topic. Oh, my gosh. Oh, my gosh. 00:00:12.300 --> 00:00:15.800 If you could think of a topic that was popular 00:00:15.800 --> 00:00:21.480 in the business world, imagine a world in which 00:00:21.480 --> 00:00:26.420 every CEO and every employee are talking about 00:00:26.420 --> 00:00:29.019 one thing, and they're talking about a different 00:00:29.019 --> 00:00:31.399 thing. Or they think they're talking about the 00:00:31.399 --> 00:00:36.840 same thing. What would that thing be? I think 00:00:36.840 --> 00:00:40.420 what they would probably be talking about is 00:00:40.420 --> 00:00:45.700 the railroad race and the golden spike of the 00:00:45.700 --> 00:00:52.859 1800s. Or perhaps any version of that over the 00:00:52.859 --> 00:00:55.520 last many centuries as you've industrialized. 00:00:56.640 --> 00:01:00.920 What I've been thinking about a lot Is is a guy 00:01:00.920 --> 00:01:03.520 named Michael Burry. Okay. Yeah, and I don't 00:01:03.520 --> 00:01:06.140 think that that's a super familiar name. Okay 00:01:06.140 --> 00:01:11.060 was folks but Author Michael Lewis made him famous 00:01:11.060 --> 00:01:15.939 in his book the big short Okay, Michael Burry 00:01:15.939 --> 00:01:21.640 is this is this savant like number cruncher that 00:01:21.640 --> 00:01:25.540 the housing market take an absolute crater Mm 00:01:25.540 --> 00:01:28.900 -hmm long before the rest of us did and that 00:01:28.900 --> 00:01:31.040 Michael Lewis attempted to explain the complex 00:01:31.040 --> 00:01:35.900 world of of collateralized debt obligations and 00:01:35.900 --> 00:01:41.140 crap mortgage and so Michael Bury made his small 00:01:41.140 --> 00:01:44.760 stack of a gazillion dollars and disappeared 00:01:44.760 --> 00:01:48.480 and so but his names turned up in the news again 00:01:48.480 --> 00:01:55.310 alongside folks like Peter Thiel as of PayPal 00:01:55.310 --> 00:02:00.909 fame and, you know, political puppeteer extraordinaire, 00:02:02.329 --> 00:02:05.310 making the world the place he wants. But these 00:02:05.310 --> 00:02:07.790 guys, their names are showing up not often together, 00:02:07.829 --> 00:02:13.110 but as individuals that are taking a look at 00:02:13.110 --> 00:02:16.430 AI and specifically some of the components of 00:02:16.430 --> 00:02:21.169 this AI boom. And they're starting to ask, at 00:02:21.169 --> 00:02:25.340 what point is this thing overheated? And so, 00:02:27.120 --> 00:02:30.080 these two individuals are both publicly recognized 00:02:30.080 --> 00:02:32.159 as guys that have started to take very short 00:02:32.159 --> 00:02:35.719 positions or selling against the companies that 00:02:35.719 --> 00:02:38.860 are so swollen in valuation and Vita being among 00:02:38.860 --> 00:02:41.800 them. And what it's made me think about, and 00:02:41.800 --> 00:02:43.800 I don't think that every employee and their employers 00:02:43.800 --> 00:02:48.099 are talking about, should we be shorting in Vita 00:02:48.099 --> 00:02:50.919 right now? But I think a lot. It's funny you 00:02:50.919 --> 00:02:53.479 call it. You say, you pronounce it differently 00:02:53.479 --> 00:02:57.620 than I. Yeah, I pronounce it invidia. Oh, invidia. 00:02:58.319 --> 00:03:00.680 I mean, I don't know if I'm right or wrong. It's 00:03:00.680 --> 00:03:03.099 probably my accent. I'm from a little further 00:03:03.099 --> 00:03:09.120 south than you are growing up down in central 00:03:09.120 --> 00:03:15.439 Idaho. Twin flats. Yeah. In the big city though, 00:03:15.479 --> 00:03:20.050 invidia. Yeah. I think yeah, I just assumed it 00:03:20.050 --> 00:03:22.789 kind of went along with in God of the Vita, you 00:03:22.789 --> 00:03:25.469 know, there you go Hey another iron button lies 00:03:25.469 --> 00:03:27.590 long. So, you know, the other the other thing 00:03:27.590 --> 00:03:30.750 that's happened this week is the Metas chief 00:03:30.750 --> 00:03:35.110 AI scientist Yan look kind look lacuna. Yeah, 00:03:35.110 --> 00:03:39.110 he plans to leave Right into the guy going to 00:03:39.110 --> 00:03:42.430 look at Bezos are probably who knows those guys 00:03:42.430 --> 00:03:45.979 just it's like horse trading at the highest level 00:03:45.979 --> 00:03:50.439 that's - Oh my gosh. Next level. So why do you 00:03:50.439 --> 00:03:55.599 think Mr. Burry's selling shorting AI stocks? 00:03:57.360 --> 00:04:01.539 I have to wonder, we've talked at different times 00:04:01.539 --> 00:04:05.259 about our career history and I remember in the 00:04:05.259 --> 00:04:09.460 late 90s and the boom of that first dot com magic, 00:04:10.819 --> 00:04:14.030 I remember thinking software, schmoftware, I 00:04:14.030 --> 00:04:15.729 want to be in the infrastructure game. I want 00:04:15.729 --> 00:04:18.769 to, I want to be where the, the demand will never 00:04:18.769 --> 00:04:21.370 slack. And so I went into it's all calm where 00:04:21.370 --> 00:04:24.149 we were laying, you know, thousands of miles 00:04:24.149 --> 00:04:27.569 of fiber optic cable just in time to watch all 00:04:27.569 --> 00:04:30.750 of this incredible investment in infrastructure 00:04:30.750 --> 00:04:34.470 implode. Yeah. And so the correlation that I 00:04:34.470 --> 00:04:37.410 can't help but observe is all this conversation 00:04:37.410 --> 00:04:41.339 about, you know, how many gallons of water and 00:04:41.339 --> 00:04:44.379 megawatts of electricity does it take when we 00:04:44.379 --> 00:04:49.500 say thank you to AI? And so the infrastructure, 00:04:49.660 --> 00:04:52.779 the swelling of it, and to bring it back to our 00:04:52.779 --> 00:04:55.959 purpose, which isn't just idle musings of two 00:04:55.959 --> 00:04:58.399 crazy middle -aged guys, but rather process. 00:04:58.879 --> 00:05:01.160 The thing that I keep thinking about, though, 00:05:01.220 --> 00:05:06.040 is every enormous leap forward in industrialization 00:05:06.040 --> 00:05:11.660 comes at a cost. And that cost is either dug 00:05:11.660 --> 00:05:15.680 out of the small pockets of a small handful of 00:05:15.680 --> 00:05:18.240 individuals, but more often out of the masses. 00:05:18.660 --> 00:05:22.279 And just like the Temple Palm crater, you know, 00:05:22.620 --> 00:05:24.959 there were huge... So take people back, because 00:05:24.959 --> 00:05:28.879 it's possible that, you know, our listeners that 00:05:28.879 --> 00:05:31.459 do not know about that crater, they may have 00:05:31.459 --> 00:05:34.620 heard about 2008 and the housing problem. the 00:05:34.620 --> 00:05:36.720 housing shortage, but they may not have known 00:05:36.720 --> 00:05:39.019 kind of what happened. Like what was the process 00:05:39.019 --> 00:05:42.040 that took place when it came to telecom? And 00:05:42.040 --> 00:05:44.839 let's tie those two together. I love it. Yep. 00:05:45.639 --> 00:05:48.819 Yeah. So for those folks that aren't turning 00:05:48.819 --> 00:05:51.920 a little bit gray around the edges, in the early 00:05:51.920 --> 00:05:56.500 2000s, right around, it was spring of 2000 exactly 00:05:56.500 --> 00:05:59.939 when this very hot dot com driven market where 00:05:59.939 --> 00:06:02.639 we couldn't put enough fiber in the ground and 00:06:02.639 --> 00:06:06.019 lay it across the oceans, All of a sudden, it 00:06:06.019 --> 00:06:09.459 caught up. The lack of capacity that drove pricing, 00:06:09.819 --> 00:06:11.620 all of a sudden, the capacity started showing 00:06:11.620 --> 00:06:15.019 up. And there was such a fervor that all of a 00:06:15.019 --> 00:06:19.540 sudden, what may have cost $10 ,000 a month in 00:06:19.540 --> 00:06:21.980 terms of subscription for a company to have this 00:06:21.980 --> 00:06:25.100 really powerful, rich access, went from 10 ,000 00:06:25.100 --> 00:06:28.800 to 1 ,000, and from 1 ,000 to 500. And today, 00:06:28.939 --> 00:06:33.819 we enjoy all kinds of insane speeds. digitally, 00:06:33.899 --> 00:06:37.000 but it's because innumerable companies went bankrupt 00:06:37.000 --> 00:06:41.699 in the early 2000s because it's so much investment 00:06:41.699 --> 00:06:45.279 in the ground that we couldn't catch up. And 00:06:45.279 --> 00:06:48.319 so a little bit of this is how many chips is 00:06:48.319 --> 00:06:53.019 enough chips? And also, can companies like Onagata 00:06:53.019 --> 00:07:00.019 DaVita or Nvidia, how much further can their 00:07:00.019 --> 00:07:04.129 valuation drive when there's a natural limit, 00:07:04.329 --> 00:07:06.790 you know, to how much production is going to 00:07:06.790 --> 00:07:09.589 take place. And this is the, I think this is 00:07:09.589 --> 00:07:11.709 the confusing part. And the thing that I saw 00:07:11.709 --> 00:07:13.850 in telecom, what was interesting was it wasn't 00:07:13.850 --> 00:07:16.689 a matter of stretching out fiber optic cables. 00:07:17.610 --> 00:07:20.550 They found was that the hardware on either end 00:07:20.550 --> 00:07:23.110 continued to get better and they got more value 00:07:23.110 --> 00:07:26.230 out of every string. And the software running 00:07:26.230 --> 00:07:28.649 on those pieces of components got increasingly 00:07:28.649 --> 00:07:32.949 complex. So let's, let's take this big picture, 00:07:32.949 --> 00:07:36.329 right? So we're back in college, right? What 00:07:36.329 --> 00:07:39.930 did we have to do to call our parents? Oh, we 00:07:39.930 --> 00:07:45.990 had an MCI calling card. You call the 1 -800 00:07:45.990 --> 00:07:49.050 number, you put in their number and it literally 00:07:49.050 --> 00:07:52.449 counted down. And this was, I literally was calling, 00:07:52.829 --> 00:07:56.750 I think maybe 40 miles away and that was long 00:07:56.750 --> 00:07:59.399 distance. Right? Yeah. And it was expensive. 00:07:59.399 --> 00:08:02.120 My grandmother lived eight miles away, growing 00:08:02.120 --> 00:08:05.019 up, and that was long distance. Yeah. And so 00:08:05.019 --> 00:08:07.300 the cost of communication was astronomical. It 00:08:07.300 --> 00:08:09.040 was super expensive. Which, what did that... 00:08:09.040 --> 00:08:11.079 Because that NCI nerd made it what, 10 cents 00:08:11.079 --> 00:08:12.800 a minute? And we thought that was smoking or 00:08:12.800 --> 00:08:14.540 maybe 15? Oh, yeah, yeah. Something like that. 00:08:14.560 --> 00:08:17.040 Like it was crazy. Because you were competing 00:08:17.040 --> 00:08:19.519 against the quarters that went into the phone 00:08:19.519 --> 00:08:24.040 machines, right? You're right. Yeah. And so the 00:08:24.040 --> 00:08:26.800 process that's leading up to this is you have... 00:08:26.860 --> 00:08:33.860 You have very, very, very profitable long -term 00:08:33.860 --> 00:08:37.039 things where you have a cliff that no one saw 00:08:37.039 --> 00:08:40.919 that they ran into it. I recently saw an AT &T 00:08:40.919 --> 00:08:44.000 commercial and it talked about how AT &T has 00:08:44.000 --> 00:08:46.039 been around since Alexander Graham Bell. And 00:08:46.039 --> 00:08:48.940 I'm like, hey, you idiots, you forget that you 00:08:48.940 --> 00:08:51.259 got bought by someone. Like you were not the 00:08:51.259 --> 00:08:53.379 buying party. You were the ones that got bought 00:08:53.379 --> 00:08:55.639 and they kept your name for marketing purposes. 00:08:55.960 --> 00:08:59.019 not because you were doing such a fantastic job. 00:09:01.100 --> 00:09:04.120 100%. That's a great call out, Chris. We may 00:09:04.120 --> 00:09:07.620 see names that sound very storied and have stood 00:09:07.620 --> 00:09:10.779 the test of time. Oftentimes, we've perpetuated 00:09:10.779 --> 00:09:13.919 those names just because they resonate. Because 00:09:13.919 --> 00:09:16.580 there's some goodwill there, which is an accounting 00:09:16.580 --> 00:09:20.759 entry somewhere. Let's bring this up to the AI. 00:09:21.769 --> 00:09:24.769 What is it with AI that Barry sees that he's 00:09:24.769 --> 00:09:26.970 like, oh my gosh, or even, you know, met his 00:09:26.970 --> 00:09:29.450 chief AI scientist is like, something's different. 00:09:29.509 --> 00:09:32.570 Like what, what are they seeing from a process 00:09:32.570 --> 00:09:36.169 standpoint that maybe we aren't, or more specifically, 00:09:36.289 --> 00:09:39.350 maybe we are seeing it, but the CEOs are not 00:09:39.350 --> 00:09:42.730 yet. Right. Well, at a certain point, if you're, 00:09:42.789 --> 00:09:45.809 if you're the one driving a train and you told 00:09:45.809 --> 00:09:49.129 the team to take the brakes out because it's 00:09:49.129 --> 00:09:52.250 slowing you down. then your only option is to 00:09:52.250 --> 00:09:57.269 go harder and go faster. Yeah, I'll confess, 00:09:57.809 --> 00:10:00.149 Burry didn't call me up to say, hey Toby, I'm 00:10:00.149 --> 00:10:03.110 taking a short position. What? I'm a little hurt, 00:10:03.149 --> 00:10:08.110 to be honest, but five bucks against some of 00:10:08.110 --> 00:10:10.509 these, that doesn't build the position that he 00:10:10.509 --> 00:10:13.269 needs to help with that choice. But I do think 00:10:13.269 --> 00:10:17.409 here's somebody who takes a very data heavy approach 00:10:17.409 --> 00:10:21.169 with this. When we start to see a number of insiders 00:10:21.169 --> 00:10:24.250 moving into stronger cash positions. I do think 00:10:24.250 --> 00:10:26.470 there's certainly a valuation question at play. 00:10:27.009 --> 00:10:29.669 But I also think again and again, we see the 00:10:29.669 --> 00:10:33.389 emergence of new technology in this mad rush 00:10:33.389 --> 00:10:37.090 to try to limp into the future. And what we're 00:10:37.090 --> 00:10:39.690 finding is that it never stacks up exactly as 00:10:39.690 --> 00:10:43.210 we thought. When informed investors start making 00:10:43.210 --> 00:10:47.590 choices that sound contrary to media, That's 00:10:47.590 --> 00:10:50.210 a really good time to take a deep breath and 00:10:50.210 --> 00:10:53.029 reassess So what's interesting is I go back to 00:10:53.029 --> 00:10:56.350 that whole tech bottle tech bubble the dot -com 00:10:56.350 --> 00:10:59.490 thing and what I go back to thinking about is 00:10:59.490 --> 00:11:04.110 like If you remember all the browsers, right? 00:11:04.250 --> 00:11:07.269 So remember and even before that there was AOL 00:11:07.269 --> 00:11:11.049 with the internet thing that just drove like 00:11:11.309 --> 00:11:14.909 Yeah, free CDs, right? Yeah. I mean, if you wanted 00:11:14.909 --> 00:11:17.889 to watch a two -minute clip, it would have taken 00:11:17.889 --> 00:11:20.970 you probably 20 minutes to download it. Like, 00:11:20.970 --> 00:11:24.570 it was incredibly slow. But the demand was there, 00:11:25.110 --> 00:11:29.509 and AOL was never gonna be dethroned. They just 00:11:29.509 --> 00:11:32.490 were gonna own the internet. They couldn't figure 00:11:32.490 --> 00:11:36.129 out how to monetize, right? And so then that 00:11:36.129 --> 00:11:38.070 basically stripped out. Like, the process there 00:11:38.070 --> 00:11:41.230 was like... at a global scale, they had to hook 00:11:41.230 --> 00:11:44.090 into that telecom thing, right? Where telecom 00:11:44.090 --> 00:11:47.350 got cheaper, right? Then other people came in 00:11:47.350 --> 00:11:51.289 and became your ISPs. And then you think about, 00:11:51.909 --> 00:11:55.870 I'm also thinking about, remember AltaVista and 00:11:55.870 --> 00:11:59.769 all these browsers, Netscape? Basically, they 00:11:59.769 --> 00:12:02.710 were really cool at first, and then they became, 00:12:03.230 --> 00:12:06.129 they couldn't figure out how to monetize. That's 00:12:06.129 --> 00:12:10.070 right. And so you ended up with ads and banner 00:12:10.070 --> 00:12:13.350 ads and everything was basically like, hey, get 00:12:13.350 --> 00:12:16.129 your attention. And then Google came out with 00:12:16.129 --> 00:12:19.590 none of that stuff and a better algorithm that 00:12:19.590 --> 00:12:25.950 cost them money. Yep. Yep. And they would have 00:12:25.950 --> 00:12:28.629 went out of business if they wouldn't have become 00:12:28.629 --> 00:12:33.250 the world's biggest advertiser. Yep. The things 00:12:33.250 --> 00:12:35.830 that are done on the front end. the advertisement 00:12:35.830 --> 00:12:38.149 where you're exposed to the monetization versus 00:12:38.149 --> 00:12:41.409 things that are done on the back end, which is 00:12:41.409 --> 00:12:43.990 harvesting the data and starting to organize 00:12:43.990 --> 00:12:48.830 it. And so we've been given the most powerful 00:12:48.830 --> 00:12:52.970 toy, I think, in our existence as a civilization, 00:12:53.129 --> 00:12:56.429 which is now AI, and people are using it to make 00:12:56.429 --> 00:13:02.080 custom Christmas lists. What if I am gluten free 00:13:02.080 --> 00:13:04.539 and replace all of the ingredients in a recipe? 00:13:05.299 --> 00:13:07.980 It's not that they're bad things. It's just that 00:13:07.980 --> 00:13:10.700 people are grappling with it and the valuations 00:13:10.700 --> 00:13:13.100 have been driven up. And so from a purely a process 00:13:13.100 --> 00:13:16.759 perspective, the speculation has gotten so far 00:13:16.759 --> 00:13:21.059 out of hand that we start to see folks that have 00:13:21.059 --> 00:13:24.820 earned a fine living, an extraordinary living, 00:13:25.399 --> 00:13:29.750 assessing human emotion and the way that Economic 00:13:29.750 --> 00:13:31.909 cycles actually move starting to make other choices. 00:13:31.909 --> 00:13:34.649 And so the good news I think for Joe consumer, 00:13:34.649 --> 00:13:38.830 you know in an office place is that The time 00:13:38.830 --> 00:13:43.309 and again these things explode. Mm -hmm On a 00:13:43.309 --> 00:13:46.570 got a devita valuation has driven all kinds of 00:13:46.570 --> 00:13:51.269 crazy Innovation, I'm sure yeah in the chip manufacturing 00:13:51.269 --> 00:13:53.269 space that we can't even guess out But everybody 00:13:53.269 --> 00:13:56.330 wants a hate of that, you know, everyone's these 00:13:56.330 --> 00:14:00.210 AI valuations Correct. And, you know, like we 00:14:00.210 --> 00:14:03.850 talked a little bit about, you know, the acronym 00:14:03.850 --> 00:14:05.850 pump, right, where it's just process. What is 00:14:05.850 --> 00:14:08.309 the process, right? And the process is, as a 00:14:08.309 --> 00:14:13.370 user, I go to a GPT or an LLM. I use that, right? 00:14:13.429 --> 00:14:15.669 What are the unintended consequences of that? 00:14:16.269 --> 00:14:18.529 Everybody's communicating the exact same way, 00:14:19.169 --> 00:14:24.740 right? Where does the money go? Not sure. That 00:14:24.740 --> 00:14:26.960 is the question right and how much money is there 00:14:26.960 --> 00:14:29.879 to go correct? and because eventually if you're 00:14:29.879 --> 00:14:32.559 investing as this as a comp as a corporation 00:14:32.559 --> 00:14:35.500 into AI you're gonna want some sort of return 00:14:35.500 --> 00:14:39.860 and generally return generally companies are 00:14:39.860 --> 00:14:42.700 wanting return on the upside on the revenue and 00:14:42.700 --> 00:14:45.580 not on not on reduction of labor costs like that 00:14:45.580 --> 00:14:51.179 is almost like exclusive Capitalism 101 is the 00:14:51.179 --> 00:14:53.740 private equity playbook right drive drive revenue 00:14:54.080 --> 00:14:57.460 Shrink expenses and just basically milk as much 00:14:57.460 --> 00:15:00.139 profit as you can well that the other piece in 00:15:00.139 --> 00:15:05.000 that Chris is is Take good hard assets and put 00:15:05.000 --> 00:15:09.240 debt on those Yeah, load up load up functioning 00:15:09.240 --> 00:15:12.480 healthy assets with debt Yep, and then pump and 00:15:12.480 --> 00:15:14.840 dump till you're right. Yeah, that playbook is 00:15:14.840 --> 00:15:16.899 an important one Yeah, yeah, and so you've got 00:15:16.899 --> 00:15:20.740 a lot of corporations that are basically I believe 00:15:21.659 --> 00:15:24.960 pumping money into AI because they believe there's 00:15:24.960 --> 00:15:28.960 an upside on the revenue and I don't see anyone 00:15:28.960 --> 00:15:33.139 making more money on AI. You know, I think I 00:15:33.139 --> 00:15:36.700 do pay $20 a month for my... I do too. Open AI 00:15:36.700 --> 00:15:39.220 subscription. Do you think that's going to be 00:15:39.220 --> 00:15:42.299 enough to sustain all this? I don't think so. 00:15:43.820 --> 00:15:47.139 I mean, it's just like Yahoo, like Yahoo, right? 00:15:49.120 --> 00:15:54.620 They were really cool, but you got to pay the 00:15:54.620 --> 00:16:00.000 piper. Yeah. It's fascinating though. And I think 00:16:00.000 --> 00:16:04.779 that as organizations, operating entities, and 00:16:04.779 --> 00:16:07.899 even as consumers, we derive a benefit. We just 00:16:07.899 --> 00:16:10.679 don't always understand the timing or depth of 00:16:10.679 --> 00:16:12.740 the benefit. And so I'll say, the trade -off 00:16:12.740 --> 00:16:16.750 from MCI cards for 10 cents a minute long distance 00:16:16.750 --> 00:16:19.429 to being able to stream entire videos on a handheld 00:16:19.429 --> 00:16:22.450 device. We're in a world that was unimaginable 00:16:22.450 --> 00:16:26.230 to us as teenagers, right? But the trade -off 00:16:26.230 --> 00:16:30.610 is that there's a world out there of information 00:16:30.610 --> 00:16:33.490 about us that we don't even know the depth of 00:16:33.490 --> 00:16:36.070 which other people are exploiting that, right? 00:16:36.750 --> 00:16:39.210 Well, and and then the reality is is like the 00:16:39.210 --> 00:16:41.129 last P of the you know, little pump after an 00:16:41.129 --> 00:16:44.750 image power right and right now the CEOs Have 00:16:44.750 --> 00:16:48.909 the power to invest Because there's still a belief 00:16:48.909 --> 00:16:53.830 in the upside of AI Yeah, and so what happens 00:16:53.830 --> 00:16:57.169 when that power? Disappears and that's that's 00:16:57.169 --> 00:17:00.090 again where we end up with this telecom bubble 00:17:00.090 --> 00:17:06.730 this This real estate bubble this Internet bubble. 00:17:06.730 --> 00:17:08.529 That's that's where you end up with these bubbles 00:17:08.529 --> 00:17:10.769 where it's like it just gets out of whack and 00:17:10.769 --> 00:17:13.609 all of a sudden Someone's got to pay the piper 00:17:13.609 --> 00:17:17.890 Yeah, and then and the other data point that 00:17:17.890 --> 00:17:19.930 I think is kind of arresting Chris along this 00:17:19.930 --> 00:17:22.130 paying the piper and the timing of infrastructure 00:17:22.130 --> 00:17:26.289 was Another I came across just maybe two weeks 00:17:26.289 --> 00:17:29.990 ago, but it was a discussion with open AI Talking 00:17:29.990 --> 00:17:32.329 with the government about downside protection 00:17:33.000 --> 00:17:35.799 Is one of the challenges well, you know all this 00:17:35.799 --> 00:17:38.440 exciting AI, but that's an enormous amount of 00:17:38.440 --> 00:17:41.599 computing power Which translates into a demand 00:17:41.599 --> 00:17:44.319 for literally thousands of data centers. Yeah, 00:17:45.140 --> 00:17:48.579 and so Here's some folks that understand the 00:17:48.579 --> 00:17:51.099 cyclical nature of this and they're saying before 00:17:51.099 --> 00:17:54.759 we commit to a gazillion dollars of data center 00:17:54.759 --> 00:17:57.539 investment We don't want to get stuck holding 00:17:57.539 --> 00:18:02.220 the bag. Yeah, it is not cheap. Oh, it's staggering 00:18:02.430 --> 00:18:08.150 Yeah, I think we'll see some change. And I think 00:18:08.150 --> 00:18:10.589 the customers will get the benefit. I think some 00:18:10.589 --> 00:18:14.210 people, some companies will end up getting themselves 00:18:14.210 --> 00:18:17.750 in big trouble. And if you think about like, 00:18:17.809 --> 00:18:21.130 oftentimes it's not the early company that wins. 00:18:22.289 --> 00:18:25.009 It's the later adopter. It's the Microsoft that 00:18:25.009 --> 00:18:29.660 comes in and kills Lotus Notes. It oftentimes 00:18:29.660 --> 00:18:33.180 it's not it's not the innovation oftentimes because 00:18:33.180 --> 00:18:36.440 they They're too busy making a market and it's 00:18:36.440 --> 00:18:38.559 the people that come in after like a Nintendo 00:18:38.559 --> 00:18:42.519 His historic like the Nintendo NES that we grew 00:18:42.519 --> 00:18:46.099 up with Mike Tyson's punch -out which find that 00:18:46.099 --> 00:18:48.259 game I would love to own it and probably terrible 00:18:48.259 --> 00:18:51.890 as terrible as I was back then But a little bit 00:18:51.890 --> 00:18:54.890 of nostalgia. Anyway, so Nintendo didn't use 00:18:54.890 --> 00:18:57.609 new tech. They used old tech and they repurposed 00:18:57.609 --> 00:18:59.250 it. And I do think that we're going to follow 00:18:59.250 --> 00:19:02.529 that same pattern where, with AI, they're going 00:19:02.529 --> 00:19:07.150 to have all of these incredibly expensive models 00:19:07.150 --> 00:19:10.349 that can do all sorts of things. But my guess 00:19:10.349 --> 00:19:14.029 is, is ChatGPT 3 .1 or 4 .1 is going to do just 00:19:14.029 --> 00:19:19.369 fine. Yeah. No, it's super interesting. Are you 00:19:19.369 --> 00:19:25.539 the blockbuster video? in 2005 or are you Netflix 00:19:25.539 --> 00:19:28.799 who you're gone all in on streaming instead of 00:19:28.799 --> 00:19:32.740 mailing physical DVDs. And so it's hard to say. 00:19:33.680 --> 00:19:35.819 We could have just as easily done this episode 00:19:35.819 --> 00:19:39.380 talking about who's bullish on AI and Jeff Betts, 00:19:39.480 --> 00:19:43.180 this brand new Prometheus startup with pocket 00:19:43.180 --> 00:19:46.500 change that he's dumped into it for like a $6 00:19:46.500 --> 00:19:50.609 billion valuation. It's a messy landscape, but 00:19:50.609 --> 00:19:54.009 I think if history is any guide, you know, all 00:19:54.009 --> 00:19:56.490 the energy and hype suggests that we're going 00:19:56.490 --> 00:19:59.549 to have some process surprises. And yet again, 00:19:59.750 --> 00:20:02.029 that pump acronym is revealing, right? Well, 00:20:02.089 --> 00:20:05.190 and the thing is, is like, it's not super clear, 00:20:05.890 --> 00:20:09.190 right? The effectiveness, the popularity is clear, 00:20:09.910 --> 00:20:15.879 but the actual getting paid and getting... getting 00:20:15.879 --> 00:20:18.880 LLMs built into your systems that add value, 00:20:19.059 --> 00:20:22.000 that actually do not require a human, that is 00:20:22.000 --> 00:20:23.900 yet to be figured out. And so maybe that's what 00:20:23.900 --> 00:20:26.660 we should talk about next time. I like it. Let's 00:20:26.660 --> 00:20:28.759 do it. All right, man. In the meantime, let me 00:20:28.759 --> 00:20:30.420 know if you have five bucks and you and I will 00:20:30.420 --> 00:20:32.900 double the short position I'm prepared to take 00:20:32.900 --> 00:20:37.599 out. Let's do it, man. Man, when you swing, you 00:20:37.599 --> 00:20:42.200 swing for the fences. All in. All right, see 00:20:42.200 --> 00:20:42.539 you.