1 00:00:02,480 --> 00:00:09,520 Speaker 1: All Zone Media. Hello, Hi, welcome to Better Offline. I'm 2 00:00:09,560 --> 00:00:23,120 Speaker 1: your host ed ze Trunk. For weeks now, I've been 3 00:00:23,120 --> 00:00:25,240 Speaker 1: writing and speaking about what I believe to be an 4 00:00:25,239 --> 00:00:29,440 Speaker 1: inevitable and possibly imminent collapse in generative AI. In the 5 00:00:29,480 --> 00:00:32,280 Speaker 1: last episode, I talked about the early signs of disquire, 6 00:00:32,640 --> 00:00:36,240 Speaker 1: particularly among investors, were growing increasingly uncomfortable about the many 7 00:00:36,240 --> 00:00:38,960 Speaker 1: billions being spent by Microsoft, Google and others to support 8 00:00:38,960 --> 00:00:41,800 Speaker 1: an industry that does not appear to make anyone much 9 00:00:41,880 --> 00:00:45,559 Speaker 1: money and costs so much more than it makes. And 10 00:00:45,640 --> 00:00:47,360 Speaker 1: I believe that this is a moment for the tech 11 00:00:47,440 --> 00:00:50,320 Speaker 1: industry to atone and to learn and to change in 12 00:00:50,320 --> 00:00:51,800 Speaker 1: the wake of what I think is an a one 13 00:00:51,920 --> 00:00:56,680 Speaker 1: hundred billion dollar group psychosis, a big, nasty waste that 14 00:00:56,800 --> 00:01:01,400 Speaker 1: I will now get to in agonizing detail. Generative AI 15 00:01:01,840 --> 00:01:05,679 Speaker 1: was always unsustainable, and it was always dependent on reams 16 00:01:05,680 --> 00:01:09,160 Speaker 1: of training data that necessitated stealing from millions of people. 17 00:01:09,360 --> 00:01:13,760 Speaker 1: With its utility vague and its ubiquity overstated, the media 18 00:01:13,800 --> 00:01:16,640 Speaker 1: and the markets have tolerated the technology that, while not 19 00:01:16,720 --> 00:01:20,320 Speaker 1: inherently bad, was implemented in a way so nefariously and 20 00:01:20,360 --> 00:01:24,440 Speaker 1: wastefully that it necessitated theft billions of dollars in cash 21 00:01:24,480 --> 00:01:28,600 Speaker 1: and the double digit percent increase in Hyperscaler's emissions. The 22 00:01:28,680 --> 00:01:32,200 Speaker 1: desperation for the tech industry to have something new has 23 00:01:32,280 --> 00:01:35,480 Speaker 1: led to such ruinous success, and if this bubble collapses, 24 00:01:35,600 --> 00:01:37,880 Speaker 1: it'll be a result of a shared myopia in both 25 00:01:37,920 --> 00:01:40,959 Speaker 1: big tech dimwitz like satch in Adela and sundarp Isshai 26 00:01:41,280 --> 00:01:44,600 Speaker 1: and Silicon Valley power players like Reid Hoffman, Sam Altman, 27 00:01:44,680 --> 00:01:48,840 Speaker 1: Bryan Chesky and of course Mark Hantreeson. The people propping 28 00:01:48,920 --> 00:01:52,160 Speaker 1: up this bubble no longer experience human problems, and thus 29 00:01:52,240 --> 00:01:56,640 Speaker 1: I don't think they can be trusted anymore to solve them. Ultimately, 30 00:01:56,760 --> 00:01:59,280 Speaker 1: this is a story of waste, ignorance and greed, of 31 00:01:59,360 --> 00:02:02,720 Speaker 1: wasted money and have wasted opportunity, because let's face it, 32 00:02:02,840 --> 00:02:05,760 Speaker 1: a fraction of the money spent by Microsoft alone chasing 33 00:02:05,800 --> 00:02:08,800 Speaker 1: the generative AI bullshit machine could have been spent in 34 00:02:08,840 --> 00:02:12,440 Speaker 1: many other profitable and societally useful ways. And it's a 35 00:02:12,480 --> 00:02:15,440 Speaker 1: story of tech executives that are so desperate to own 36 00:02:15,480 --> 00:02:19,320 Speaker 1: the future, but are so disconnected from actually building it. 37 00:02:20,320 --> 00:02:22,480 Speaker 1: And this arms race is a monument to the lack 38 00:02:22,520 --> 00:02:25,760 Speaker 1: of curiosity rife in the highest ranks of the tech industry. 39 00:02:26,400 --> 00:02:28,800 Speaker 1: They refuse to do the hard work to create, to 40 00:02:28,800 --> 00:02:31,480 Speaker 1: be curious, to be excited about the things they're building 41 00:02:31,600 --> 00:02:34,360 Speaker 1: and the people they're serving, And so they spent billions 42 00:02:34,400 --> 00:02:36,680 Speaker 1: of dollars to eliminate the risk they even might have 43 00:02:36,760 --> 00:02:39,440 Speaker 1: to do any of those things. Ever, again, they just 44 00:02:39,520 --> 00:02:42,560 Speaker 1: wanted the machine to do it for them. Had Sandhar 45 00:02:42,600 --> 00:02:45,400 Speaker 1: Pishai looked at Microsoft's investment in open Ai in twenty 46 00:02:45,520 --> 00:02:47,880 Speaker 1: nineteen and said, nah, no, thanks, which you did with 47 00:02:47,919 --> 00:02:50,360 Speaker 1: the metavers. By the way, it's likely that none of 48 00:02:50,360 --> 00:02:53,400 Speaker 1: this would have happened. But a combined hunger for growth 49 00:02:53,440 --> 00:02:55,600 Speaker 1: and a lack of any natural predators means that big 50 00:02:55,639 --> 00:02:58,000 Speaker 1: tech no longer knows how to make a competitive or 51 00:02:58,080 --> 00:03:00,440 Speaker 1: useful product, and thus they can only see what their 52 00:03:00,440 --> 00:03:03,320 Speaker 1: competitors are doing and say, ah, yeah, yeah, that's what 53 00:03:03,360 --> 00:03:06,320 Speaker 1: it is. That's what we're all doing now. Before the 54 00:03:06,320 --> 00:03:09,760 Speaker 1: AI hype boom, Mark Zuckerberg was once so disconnected from 55 00:03:09,760 --> 00:03:12,600 Speaker 1: Meta's work on AI that he literally had no idea 56 00:03:12,680 --> 00:03:15,760 Speaker 1: what breakthrough Sandar Peshai had complimented him on in a 57 00:03:15,800 --> 00:03:18,520 Speaker 1: meeting they had together mere months before. Meta's own obsession 58 00:03:18,560 --> 00:03:22,120 Speaker 1: with AI truly began, and none of these guys seem 59 00:03:22,200 --> 00:03:24,200 Speaker 1: to have any idea what's going on and why are 60 00:03:24,200 --> 00:03:27,280 Speaker 1: they having these chummy little meetings. These aren't competitors, they're 61 00:03:27,280 --> 00:03:30,680 Speaker 1: co conspirators. They're just people meeting to say, what's the 62 00:03:30,800 --> 00:03:34,120 Speaker 1: latest con we're all jumping on. What's the latest market 63 00:03:34,240 --> 00:03:36,680 Speaker 1: that we're all going to monopolize. I can't wait for 64 00:03:36,720 --> 00:03:40,240 Speaker 1: the monopoly episode. And these companies, they're too large, they 65 00:03:40,240 --> 00:03:43,280 Speaker 1: are too unwieldy, they're too disconnected, and they do too much. 66 00:03:43,760 --> 00:03:46,760 Speaker 1: They lack the focus that makes a truly competitive business, 67 00:03:46,920 --> 00:03:49,680 Speaker 1: and they lack any kind of cohesive culture, the kinds 68 00:03:49,680 --> 00:03:52,440 Speaker 1: that are built on solving real human problems or real 69 00:03:52,520 --> 00:03:56,080 Speaker 1: business problems that humans have. These are not companies built 70 00:03:56,120 --> 00:03:59,640 Speaker 1: for anything other than growth. And none of them, not 71 00:03:59,680 --> 00:04:03,240 Speaker 1: even Apple, have built something truly innovative and life changing 72 00:04:03,240 --> 00:04:05,640 Speaker 1: in the best part of a decade, with the exception 73 00:04:05,840 --> 00:04:09,840 Speaker 1: perhaps of maybe contactless mobile payments, which are actually really cool. 74 00:04:10,600 --> 00:04:12,880 Speaker 1: These companies are run by rot economists, and they have 75 00:04:12,920 --> 00:04:16,320 Speaker 1: these disconnected, chaotic cultures full of these little fiefdoms. We're 76 00:04:16,440 --> 00:04:19,320 Speaker 1: established technologists are rap fuck by management goons when they 77 00:04:19,360 --> 00:04:22,200 Speaker 1: try to make their products better instead of making them 78 00:04:22,200 --> 00:04:25,640 Speaker 1: more money, and the depressing thing is that they really 79 00:04:25,760 --> 00:04:29,200 Speaker 1: don't need to be that way. Seriously, there's a world 80 00:04:29,240 --> 00:04:32,039 Speaker 1: where these companies just make a billion dollars a quarter 81 00:04:32,400 --> 00:04:34,520 Speaker 1: and they don't even have to fire people every time 82 00:04:34,560 --> 00:04:37,440 Speaker 1: they do so. And there's one where these companies actually 83 00:04:37,440 --> 00:04:40,360 Speaker 1: solve real problems too, and then they still make incredibly 84 00:04:40,440 --> 00:04:43,760 Speaker 1: large amounts of money for doing so. The problem is 85 00:04:43,800 --> 00:04:46,960 Speaker 1: they're greedy. They run by greedy idiots, and they're addicted 86 00:04:47,040 --> 00:04:49,400 Speaker 1: to growth and are incapable of doing anything other than 87 00:04:49,440 --> 00:04:52,880 Speaker 1: following what the last guy did. And what's funny is 88 00:04:53,320 --> 00:04:57,440 Speaker 1: this is very Jack Welch from the Shareholder Supremacy. They 89 00:04:57,480 --> 00:04:59,400 Speaker 1: just see a market and like, what if we put 90 00:04:59,440 --> 00:05:01,640 Speaker 1: money in that? Can we take the money from the market. 91 00:05:01,720 --> 00:05:05,240 Speaker 1: That's all they can do. Now. Sundarpishi is a terrible executive, 92 00:05:05,400 --> 00:05:08,279 Speaker 1: but he's a really good Jack Welch clone. And by 93 00:05:08,320 --> 00:05:09,720 Speaker 1: the way, if you don't know who Jack Welch is, 94 00:05:09,760 --> 00:05:13,360 Speaker 1: please go back to the Shareholder Supremacy episode. It's important 95 00:05:13,760 --> 00:05:19,520 Speaker 1: because you really need to see who fucked all this up. Okay, 96 00:05:19,839 --> 00:05:24,360 Speaker 1: so where were we what happens next? Well, there's enough 97 00:05:24,360 --> 00:05:26,800 Speaker 1: money sloshing around right now that open ai probably can 98 00:05:26,839 --> 00:05:29,919 Speaker 1: be bankrolled a little further, and like I've mentioned in 99 00:05:29,960 --> 00:05:32,000 Speaker 1: the previous episode, it'll need to be is it needs 100 00:05:32,000 --> 00:05:34,760 Speaker 1: another five billion dollars, if not ten billion dollars to 101 00:05:34,760 --> 00:05:37,680 Speaker 1: make it to the end of twenty twenty five, That 102 00:05:37,839 --> 00:05:39,600 Speaker 1: very large amount of money's going to have to come 103 00:05:39,600 --> 00:05:41,640 Speaker 1: from somewhere, and I think that they're going to be 104 00:05:41,680 --> 00:05:43,760 Speaker 1: reduced to go into the Kingdom of Saudi Arabia or 105 00:05:43,800 --> 00:05:46,920 Speaker 1: soft bank, or probably taking on private debt, and that 106 00:05:47,000 --> 00:05:49,440 Speaker 1: last one, I'm just really not sure is going to happen. 107 00:05:50,120 --> 00:05:51,880 Speaker 1: And those suitors they're not going to be quite as 108 00:05:51,920 --> 00:05:54,360 Speaker 1: generous with their terms and they won't give them a 109 00:05:54,400 --> 00:05:57,680 Speaker 1: devil's deal like Microsoft did, which, as I previously argued 110 00:05:57,720 --> 00:06:01,600 Speaker 1: across multiple newsletters and podcasts, effectively owns all of open 111 00:06:01,640 --> 00:06:05,960 Speaker 1: AI's research and intellectual property. Someonlman is a terrible deal maker. 112 00:06:06,000 --> 00:06:09,360 Speaker 1: By the way, while the venture capital system theoretically could 113 00:06:09,360 --> 00:06:11,920 Speaker 1: afford to prop up open ai for another year or so, 114 00:06:12,279 --> 00:06:15,560 Speaker 1: they're limited partners. Who are the people funding the vcs. 115 00:06:16,080 --> 00:06:18,200 Speaker 1: They're probably still a little sore from twenty twenty one 116 00:06:18,240 --> 00:06:20,560 Speaker 1: and twenty twenty two, when massive amounts of money went 117 00:06:20,600 --> 00:06:23,360 Speaker 1: into nothing, not to mention the sudden and rapid rise 118 00:06:23,400 --> 00:06:26,240 Speaker 1: and interest rates, and they'll probably bulk at the amounts 119 00:06:26,240 --> 00:06:30,799 Speaker 1: they can need. We're talking multiple, probably ten one hundred 120 00:06:30,839 --> 00:06:33,559 Speaker 1: million dollar checks or two hundred million dollar checks. Open 121 00:06:33,600 --> 00:06:36,440 Speaker 1: ai needs so much more money than the VC system 122 00:06:36,480 --> 00:06:39,520 Speaker 1: is ready to give to one company, and it really 123 00:06:39,520 --> 00:06:42,080 Speaker 1: will only take one person blinking to create a panic 124 00:06:42,120 --> 00:06:45,039 Speaker 1: in both the public and the private markets. Open Ai 125 00:06:45,200 --> 00:06:47,680 Speaker 1: may still be able to raise money, but investors have 126 00:06:47,720 --> 00:06:51,640 Speaker 1: no real path to liquidity as open ai most assuredly 127 00:06:51,680 --> 00:06:53,839 Speaker 1: will not be able to go public, and in the 128 00:06:53,880 --> 00:06:57,400 Speaker 1: process of investing these vcs will likely find out exactly 129 00:06:57,400 --> 00:07:00,000 Speaker 1: how bad things are there. While there is a chance 130 00:07:00,120 --> 00:07:02,160 Speaker 1: we get one more round, kind of a dead cat 131 00:07:02,200 --> 00:07:05,359 Speaker 1: bounce situation, it's honestly a little tough to imagine how 132 00:07:05,440 --> 00:07:08,160 Speaker 1: it keeps going even if open ai raises another ten 133 00:07:08,200 --> 00:07:13,040 Speaker 1: billion dollars. But how does the collapse happen? Well, as 134 00:07:13,040 --> 00:07:15,400 Speaker 1: I said last time, there's the obvious one open ai 135 00:07:15,480 --> 00:07:18,560 Speaker 1: or Anthropic shuts down, but I think that them shutting 136 00:07:18,680 --> 00:07:22,239 Speaker 1: down is very unlikely based on the corporate structure open 137 00:07:22,280 --> 00:07:25,040 Speaker 1: AI or Anthropic collapsing will be more like the character 138 00:07:25,040 --> 00:07:28,160 Speaker 1: dot Ai deal an attempt to hide the ugly mess, 139 00:07:28,160 --> 00:07:30,320 Speaker 1: where a big tech firm absorbs them in and just 140 00:07:30,360 --> 00:07:32,800 Speaker 1: goes there. Now the AI division. You saw a version 141 00:07:32,800 --> 00:07:34,960 Speaker 1: of this, by the way, when Sam got fired from 142 00:07:35,000 --> 00:07:37,360 Speaker 1: open Ai and he was like the head of AI 143 00:07:37,480 --> 00:07:40,320 Speaker 1: at Microsoft for two minutes. But if I had to guess, 144 00:07:40,360 --> 00:07:43,280 Speaker 1: Anthropics most likely home would be with Amazon, which completed 145 00:07:43,280 --> 00:07:45,720 Speaker 1: its four billion dollar investment into the company earlier this 146 00:07:45,840 --> 00:07:49,000 Speaker 1: year and also has a profit share with them. In 147 00:07:49,080 --> 00:07:51,960 Speaker 1: any case where this happens, even if the entity Anthropical 148 00:07:52,040 --> 00:07:54,240 Speaker 1: or open Ai keeps going without the key people in 149 00:07:54,320 --> 00:07:57,600 Speaker 1: power such as Anthropic CEO Dario Omdey going in house 150 00:07:57,680 --> 00:08:02,680 Speaker 1: Amazon or Google again, it doesn't really matter. They will 151 00:08:02,800 --> 00:08:07,000 Speaker 1: eventually die and raising around for a generative AI company 152 00:08:07,000 --> 00:08:08,960 Speaker 1: in this market, it's likely going to be quite difficult, 153 00:08:09,000 --> 00:08:12,400 Speaker 1: if not impossible. The narrative shifted, and the people with 154 00:08:12,440 --> 00:08:14,480 Speaker 1: the money to invest don't really know what they're talking 155 00:08:14,520 --> 00:08:18,240 Speaker 1: about technologically speaking, so they're investing based on how everybody 156 00:08:18,280 --> 00:08:20,440 Speaker 1: else is feeling, what the vibes are in the air, 157 00:08:21,080 --> 00:08:23,640 Speaker 1: what they sniff each other's asses, and they say, oh, fuck, 158 00:08:23,720 --> 00:08:25,680 Speaker 1: is AI good or bad? Now? I don't really know, 159 00:08:26,680 --> 00:08:30,000 Speaker 1: And when we really strip away everything. We strip away 160 00:08:30,040 --> 00:08:33,040 Speaker 1: all of the lies and the marketing, and you just 161 00:08:33,080 --> 00:08:36,240 Speaker 1: get down to incentives and you think, well, they just 162 00:08:36,280 --> 00:08:39,800 Speaker 1: want money, right, If that's all they want, that really 163 00:08:39,880 --> 00:08:44,000 Speaker 1: is just the problem, because fundamentally this tech really isn't 164 00:08:44,040 --> 00:08:48,280 Speaker 1: that exciting. I'm sorry it isn't. Generative AI is not 165 00:08:48,400 --> 00:08:51,679 Speaker 1: the AI from Iron Man. It is not the magical 166 00:08:51,720 --> 00:08:56,119 Speaker 1: future computer. It's a generative AI. That picture of Garfield 167 00:08:56,120 --> 00:08:58,560 Speaker 1: with a gun or Scooby Doo sitting on top of 168 00:08:58,559 --> 00:09:03,440 Speaker 1: a rocket, those are aren't useful. That's not the future. 169 00:09:04,160 --> 00:09:08,120 Speaker 1: That's just image creation. I'm sure some very insufferable freak 170 00:09:08,200 --> 00:09:09,800 Speaker 1: is going to email me after this and be like, wow, 171 00:09:09,840 --> 00:09:13,960 Speaker 1: I'm set up generative AI in all of these ways. Mate. 172 00:09:14,120 --> 00:09:16,640 Speaker 1: I appreciate your time, Thank you for emailing me. I'm 173 00:09:16,640 --> 00:09:19,200 Speaker 1: sure you don't even sound like that. The point I'm 174 00:09:19,200 --> 00:09:22,520 Speaker 1: not making is that this is useless. The point I'm 175 00:09:22,520 --> 00:09:25,439 Speaker 1: making is it is not even in the same galaxy 176 00:09:25,559 --> 00:09:29,080 Speaker 1: of usefulness that it requires to go forward to be 177 00:09:29,160 --> 00:09:31,199 Speaker 1: worth all of this money. And on top of that, 178 00:09:31,840 --> 00:09:36,360 Speaker 1: it's boring. It's so goddamn boring. It's so I am 179 00:09:36,520 --> 00:09:40,120 Speaker 1: so bored. It doesn't even do my boring business stuff. 180 00:09:40,200 --> 00:09:42,000 Speaker 1: I have a pr business in my day job, but 181 00:09:42,000 --> 00:09:44,480 Speaker 1: I would love a dn ky AI to move my 182 00:09:44,520 --> 00:09:46,880 Speaker 1: shit around. It can't even do that. I want it 183 00:09:46,960 --> 00:09:49,120 Speaker 1: to help me, and it won't do it. And if 184 00:09:49,160 --> 00:09:51,880 Speaker 1: you say ed, you're a stupid idiot, You're a moron. 185 00:09:51,960 --> 00:09:53,440 Speaker 1: You have the brain of a bird or a dog 186 00:09:53,559 --> 00:09:56,560 Speaker 1: or a dog bird hybrid. You drool all the time 187 00:09:56,640 --> 00:09:59,480 Speaker 1: because your brain cannot keep your mouth closed. All of 188 00:09:59,480 --> 00:10:02,040 Speaker 1: these things are true, but it's not my job to 189 00:10:02,080 --> 00:10:06,200 Speaker 1: sell Generative AI to myself. It's the companies. And the 190 00:10:06,240 --> 00:10:08,839 Speaker 1: idea of a super smart friend that knows everything about 191 00:10:08,840 --> 00:10:11,120 Speaker 1: me is exciting, don't get me wrong. As is the 192 00:10:11,160 --> 00:10:14,840 Speaker 1: idea of making away drudgery all of my spreadsheets and documents, 193 00:10:14,880 --> 00:10:18,080 Speaker 1: for example. But Generative AI does not do these things, 194 00:10:18,440 --> 00:10:21,760 Speaker 1: putting aside any personal feelings I may have. It has 195 00:10:21,800 --> 00:10:24,520 Speaker 1: not changed my life or even like slightly improved it. 196 00:10:24,679 --> 00:10:26,400 Speaker 1: And I'm a guy who loves using tech, and he 197 00:10:26,440 --> 00:10:29,600 Speaker 1: is willing to put hours hours into finding solutions to 198 00:10:29,640 --> 00:10:33,480 Speaker 1: automate or mitigate very small problems that I only really 199 00:10:33,640 --> 00:10:37,160 Speaker 1: deal with occasionally. I'm not lazy, I'm efficient, at least 200 00:10:37,200 --> 00:10:39,800 Speaker 1: that's what I tell my doctor, and I don't know 201 00:10:39,840 --> 00:10:42,600 Speaker 1: a single person that has meaningfully changed their workflow or 202 00:10:42,640 --> 00:10:45,360 Speaker 1: life as a result of generative AI, outside of being 203 00:10:45,400 --> 00:10:47,480 Speaker 1: able to integrate it to speed up some part of 204 00:10:47,520 --> 00:10:52,360 Speaker 1: their already existing system. I've literally never met or spoken 205 00:10:52,400 --> 00:10:56,240 Speaker 1: to someone other than one guy on Twitter who has 206 00:10:56,280 --> 00:10:58,720 Speaker 1: talked about using this every day in a way that 207 00:10:58,800 --> 00:11:02,920 Speaker 1: truly changes things, and I should. As an aside, I 208 00:11:02,960 --> 00:11:05,600 Speaker 1: do occasionally get someone on Twitter and it's always a 209 00:11:05,600 --> 00:11:08,040 Speaker 1: blue check and they always say, oh, yeah, well, it's 210 00:11:08,080 --> 00:11:10,760 Speaker 1: changed the way our code changy allows me to code 211 00:11:10,800 --> 00:11:14,439 Speaker 1: so much faster. They then get eighteen responses of someone saying, 212 00:11:14,480 --> 00:11:19,280 Speaker 1: I'm an engineer and this is very unstable, and it 213 00:11:19,360 --> 00:11:23,200 Speaker 1: really feels like generative AI is just catnip for the 214 00:11:23,360 --> 00:11:26,360 Speaker 1: truly incurious. I don't want to say lazy, because I 215 00:11:26,360 --> 00:11:28,800 Speaker 1: think everyone is kind of lazy on some level. We 216 00:11:28,880 --> 00:11:31,560 Speaker 1: all want things to be easier, we really do. And 217 00:11:31,600 --> 00:11:34,800 Speaker 1: there is, when it comes to creativity, a degree where 218 00:11:36,200 --> 00:11:39,200 Speaker 1: the problems you face making something are what makes it 219 00:11:39,240 --> 00:11:42,080 Speaker 1: what it is. The struggles I've had writing and reading 220 00:11:42,120 --> 00:11:44,559 Speaker 1: this script are what make me a better orator or 221 00:11:44,600 --> 00:11:48,000 Speaker 1: a better writer. That process is important, but it's hard, 222 00:11:48,040 --> 00:11:50,520 Speaker 1: and it's annoying, and you have to suck a few times, 223 00:11:50,720 --> 00:11:53,480 Speaker 1: you have to write crap, You have to retake things 224 00:11:53,480 --> 00:11:56,400 Speaker 1: three or four times, and your poor editor, Mattasowski sitting 225 00:11:56,400 --> 00:11:58,760 Speaker 1: there going, mate, you've reread that paragraph three times, you 226 00:11:58,840 --> 00:12:02,240 Speaker 1: messed the same word up. But it's fun. This is 227 00:12:02,240 --> 00:12:05,360 Speaker 1: what curiosity is. And I don't think these people have it. 228 00:12:05,520 --> 00:12:07,839 Speaker 1: These people don't want to solve software problems. They want 229 00:12:07,840 --> 00:12:10,480 Speaker 1: to type a thing into a computer, hit enter and 230 00:12:10,520 --> 00:12:13,679 Speaker 1: then the problem is solved, never understanding why they did it. 231 00:12:14,280 --> 00:12:16,680 Speaker 1: And this isn't the same thing as calculators, by the way, 232 00:12:17,160 --> 00:12:19,520 Speaker 1: I'd argue software development as are, but that's a separate thing. 233 00:12:20,679 --> 00:12:25,240 Speaker 1: And ultimately, this whole movement just doesn't even feel real. 234 00:12:25,880 --> 00:12:29,280 Speaker 1: Nothing about it really felt real. It's just a farce. 235 00:12:29,960 --> 00:12:32,240 Speaker 1: It costs so much money, and it's still costing so 236 00:12:32,320 --> 00:12:35,960 Speaker 1: much money today, and the promises of artificial intelligence will 237 00:12:35,960 --> 00:12:38,959 Speaker 1: not be kept through the pursuit of generative AI. As 238 00:12:39,000 --> 00:12:41,720 Speaker 1: generative AI is a dead end technology that has peaked, 239 00:12:41,960 --> 00:12:44,559 Speaker 1: and it costs too much both financially and socially to 240 00:12:44,640 --> 00:12:48,400 Speaker 1: keep it going or to justify its continued existence. There's 241 00:12:48,400 --> 00:12:51,600 Speaker 1: no heroic story here. There are a few people who 242 00:12:51,679 --> 00:12:54,720 Speaker 1: are sincerely cheering for these companies to win, those who, 243 00:12:54,760 --> 00:12:57,680 Speaker 1: of course don't have a tangible financial term banking on 244 00:12:57,720 --> 00:13:00,280 Speaker 1: the value of open AI or they have so one 245 00:13:00,320 --> 00:13:03,120 Speaker 1: of the one hundred different guys claiming the Heaven AI newsletter, 246 00:13:03,559 --> 00:13:06,960 Speaker 1: and let's be honest, nobody really loves this, do they. 247 00:13:07,080 --> 00:13:10,320 Speaker 1: Nobody really uses CHATGBT and has that oh shit moment 248 00:13:10,320 --> 00:13:12,520 Speaker 1: you got when you first use an iPhone, when you're 249 00:13:12,520 --> 00:13:16,200 Speaker 1: first sent a file with Dropbox, that those little times 250 00:13:16,240 --> 00:13:18,880 Speaker 1: when you really fell in love with technology again and again. 251 00:13:19,840 --> 00:13:22,040 Speaker 1: And there's no magic in it. There's no whimsy, there's 252 00:13:22,080 --> 00:13:25,679 Speaker 1: no joy in any of these larger language models, and 253 00:13:25,720 --> 00:13:29,400 Speaker 1: they're not really helping enough people to make it worthwhile. 254 00:13:29,960 --> 00:13:33,560 Speaker 1: I nearly said they're not helping anybody. But there are users, 255 00:13:33,920 --> 00:13:37,240 Speaker 1: there are people who actually care about this. How much 256 00:13:37,280 --> 00:13:40,320 Speaker 1: do they really care? And does any of that outweigh 257 00:13:40,320 --> 00:13:43,360 Speaker 1: the costs and the environmental damage. The answer is no, 258 00:13:43,480 --> 00:13:46,960 Speaker 1: by the way, and their peddlers always demand that we 259 00:13:47,160 --> 00:13:49,760 Speaker 1: just give them another minute as they burn billions of 260 00:13:49,800 --> 00:13:53,080 Speaker 1: dollars to get nowhere. And the general public has known 261 00:13:53,160 --> 00:13:56,040 Speaker 1: the problems and limitations of generative AI for months. I 262 00:13:56,160 --> 00:14:01,520 Speaker 1: constantly get emails from better offline listeners, artists, engineers, teachers, massionists, 263 00:14:01,559 --> 00:14:05,280 Speaker 1: poison makers, journalists, consultants, cleaners, and copy editors, all saying 264 00:14:05,280 --> 00:14:07,880 Speaker 1: the same thing. This all seems like a croc of shit, 265 00:14:08,040 --> 00:14:11,040 Speaker 1: and the outputs are terrible because people who actually make 266 00:14:11,120 --> 00:14:14,959 Speaker 1: things respect the output and the production itself, and they 267 00:14:14,960 --> 00:14:18,920 Speaker 1: can see how bad this stuff is. Generative AI was 268 00:14:18,960 --> 00:14:21,920 Speaker 1: also not peddled as a solution to any problem other 269 00:14:21,960 --> 00:14:24,640 Speaker 1: than tech's lack of a hypergrowth market, and it's the 270 00:14:24,720 --> 00:14:27,120 Speaker 1: kind of bet that only a tech industry devoid of 271 00:14:27,200 --> 00:14:31,120 Speaker 1: natural selection truly builds. None of these big tech firms, 272 00:14:31,160 --> 00:14:34,800 Speaker 1: and I'd argue venture capitalists really take risks anymore because 273 00:14:34,800 --> 00:14:36,880 Speaker 1: they don't want to end up looking stupid, and they 274 00:14:36,880 --> 00:14:40,040 Speaker 1: want to follow each other. And regular people know. They 275 00:14:40,080 --> 00:14:43,200 Speaker 1: know Google is worse, they know Facebook is bordering unusable, 276 00:14:43,400 --> 00:14:46,520 Speaker 1: They know that generative AI does not do anything obviously 277 00:14:46,600 --> 00:14:49,760 Speaker 1: useful at scale, and that these companies have never made 278 00:14:49,760 --> 00:14:52,160 Speaker 1: more money than they're making today as they fuck the 279 00:14:52,160 --> 00:14:54,960 Speaker 1: customer again and again. And I believe we're in the 280 00:14:55,000 --> 00:14:57,280 Speaker 1: beginnings of a true revolt against big tech and their 281 00:14:57,320 --> 00:15:00,800 Speaker 1: rotten little promises, their lucks run out, and these formless 282 00:15:00,800 --> 00:15:03,480 Speaker 1: private equity vehicles dressed up as software companies are no 283 00:15:03,560 --> 00:15:06,840 Speaker 1: longer bred to create new things. We're all sitting and 284 00:15:06,880 --> 00:15:09,920 Speaker 1: staring at innovation's corps, stabbed full of holes by McKinsey 285 00:15:10,000 --> 00:15:12,960 Speaker 1: veterans and slimy con artists, and it's time for men 286 00:15:13,040 --> 00:15:16,520 Speaker 1: like Sandar Pashai and Sachin Adela to atone because their 287 00:15:16,520 --> 00:15:19,440 Speaker 1: big bet looks like a gigantic waste, one so obviously 288 00:15:19,520 --> 00:15:21,680 Speaker 1: rotten from the beginning that they've never been able to 289 00:15:21,680 --> 00:15:25,080 Speaker 1: tell anyone that they never had a plan beyond throw 290 00:15:25,200 --> 00:15:27,280 Speaker 1: money at it and eventually it will do something, or 291 00:15:27,440 --> 00:15:29,200 Speaker 1: we will do this as long as the other guy 292 00:15:29,320 --> 00:15:32,960 Speaker 1: does it too. This is a big, ugly mess, a 293 00:15:33,040 --> 00:15:35,480 Speaker 1: monument to the excess that burned more than two hundred 294 00:15:35,520 --> 00:15:39,120 Speaker 1: billion dollars and allowed equally lazy and uncreative people to 295 00:15:39,160 --> 00:15:42,760 Speaker 1: replace real people with shitty facsimiles that are just good enough. 296 00:15:43,880 --> 00:15:45,640 Speaker 1: But do you want to be more than good enough? 297 00:15:45,920 --> 00:15:47,560 Speaker 1: Do you want to excel in all things as I 298 00:15:47,680 --> 00:15:50,320 Speaker 1: demand of all my listeners, then buy one of the 299 00:15:50,320 --> 00:15:54,560 Speaker 1: following products, which, as usual, will reflect my exact beliefs, 300 00:15:54,840 --> 00:16:09,320 Speaker 1: wants and needs and we're back, and now that I'm back, 301 00:16:09,600 --> 00:16:11,400 Speaker 1: I want to say something to my peers in the media. 302 00:16:12,120 --> 00:16:14,680 Speaker 1: It's time for a fundamental re evaluation of how we 303 00:16:14,760 --> 00:16:17,480 Speaker 1: treat big tech, and it's time to question the promises 304 00:16:17,520 --> 00:16:20,880 Speaker 1: of those who sought to peddle this generative nonsense. It's 305 00:16:20,960 --> 00:16:23,840 Speaker 1: no longer time to ask Mark Zuckerberg about his chain. 306 00:16:24,240 --> 00:16:26,160 Speaker 1: It's time to ask him how long he intends to 307 00:16:26,160 --> 00:16:28,600 Speaker 1: burn tens of billions of dollars on nothing, and why 308 00:16:28,600 --> 00:16:30,840 Speaker 1: the website that made him so many billions of dollars 309 00:16:30,880 --> 00:16:33,120 Speaker 1: is falling apart at the seams, pumped full of content 310 00:16:33,360 --> 00:16:37,080 Speaker 1: created by the big, stupid bullshit machine he's helped popularize. 311 00:16:37,440 --> 00:16:40,880 Speaker 1: It's no longer time to gingerly cool about Sundar pishe Oh, 312 00:16:40,920 --> 00:16:43,760 Speaker 1: he's looking at my smartphone. It's time to ask him 313 00:16:43,760 --> 00:16:45,800 Speaker 1: how Gemini is going to make a profit and what 314 00:16:45,920 --> 00:16:47,840 Speaker 1: it is that it will do, and press him again 315 00:16:47,880 --> 00:16:49,920 Speaker 1: and again and again and again and again until he 316 00:16:49,960 --> 00:16:53,520 Speaker 1: talks about why Google Search is so bad. Now, it's 317 00:16:53,560 --> 00:16:56,240 Speaker 1: no longer time to let Sach and Adela oscillate from 318 00:16:56,280 --> 00:16:59,600 Speaker 1: referendums on capitalism twenty twenty to laying off tens of 319 00:16:59,600 --> 00:17:02,200 Speaker 1: thousands of people in twenty twenty three. It's time to 320 00:17:02,280 --> 00:17:05,040 Speaker 1: ask him if he's prepared to bankroll open AI to 321 00:17:05,080 --> 00:17:07,520 Speaker 1: the tune of thirty billion dollars and refuse to leave 322 00:17:07,560 --> 00:17:11,840 Speaker 1: the conversation without committing to a number. These men have 323 00:17:11,920 --> 00:17:14,399 Speaker 1: burned hundreds of billions of dollars on a machine that 324 00:17:14,440 --> 00:17:17,159 Speaker 1: boils lakes to make the most mediocre version of the past. 325 00:17:18,000 --> 00:17:21,040 Speaker 1: These men are not supporting innovation. They're aggressive forces, too 326 00:17:21,080 --> 00:17:23,439 Speaker 1: scared to piss off the markets to actually change the world, 327 00:17:23,600 --> 00:17:26,439 Speaker 1: domesticated animals that fear no natural predators and thus have 328 00:17:26,560 --> 00:17:29,600 Speaker 1: no hunger for something new. They propped up generative AI 329 00:17:29,680 --> 00:17:31,960 Speaker 1: because their followers rather than leaders that have paid hundreds 330 00:17:32,000 --> 00:17:34,439 Speaker 1: of millions of dollars to empower a nihilistic kind of 331 00:17:34,440 --> 00:17:38,560 Speaker 1: growth capitalism that dresses up monopolies as disruption. They have 332 00:17:38,640 --> 00:17:42,000 Speaker 1: propped up generative AI because building the future is hard 333 00:17:42,160 --> 00:17:45,119 Speaker 1: and expensive and involves taking risks and making sure that 334 00:17:45,160 --> 00:17:47,159 Speaker 1: you have a culture that doesn't expect the answer to 335 00:17:47,200 --> 00:17:50,400 Speaker 1: a question in the fastest amount of time possible. Innovation 336 00:17:50,480 --> 00:17:53,720 Speaker 1: is an efficient. It's lossy. It has indeterminate time scales 337 00:17:53,760 --> 00:17:56,920 Speaker 1: and requires both capital investment and space required for great 338 00:17:56,960 --> 00:17:59,679 Speaker 1: people to think of actual solutions to problems that involve 339 00:17:59,720 --> 00:18:03,359 Speaker 1: them talking to and caring about the lives of real people. 340 00:18:04,720 --> 00:18:08,040 Speaker 1: Fire satch In the Dellar of Microsoft, Fire sund Arpishai 341 00:18:08,119 --> 00:18:10,720 Speaker 1: of Google, and work out a way to fire Mark 342 00:18:10,800 --> 00:18:14,919 Speaker 1: Zuckerberg of Meta. These men are unworthy. They do not 343 00:18:15,160 --> 00:18:19,200 Speaker 1: find any of this exciting or meaningful. They have no curiosity. 344 00:18:19,320 --> 00:18:22,800 Speaker 1: They're private equity guys dressed up as software guys. They 345 00:18:22,800 --> 00:18:26,160 Speaker 1: are con artists. They are running their products into the ground, 346 00:18:26,600 --> 00:18:30,000 Speaker 1: and they're making more money than ever. Big Tech has 347 00:18:30,000 --> 00:18:33,480 Speaker 1: been overcome with narcissism and nihilism, and it must change 348 00:18:33,480 --> 00:18:37,480 Speaker 1: its ways to not eventually face perdition. I am fucking serious. 349 00:18:38,080 --> 00:18:40,320 Speaker 1: Do you think these people have a backup plan? Do 350 00:18:40,320 --> 00:18:42,160 Speaker 1: you think they've burned all of this money on such 351 00:18:42,160 --> 00:18:46,280 Speaker 1: a mediocre and wasteful solution because they've got other better ideas? No, 352 00:18:46,720 --> 00:18:49,800 Speaker 1: they're all out. Big Tech does not have another thing. 353 00:18:50,080 --> 00:18:53,199 Speaker 1: Google doesn't have any ideas. Meta is so out of 354 00:18:53,280 --> 00:18:56,159 Speaker 1: ideas that they change their name to another worse idea, 355 00:18:56,320 --> 00:18:59,639 Speaker 1: one associated with Mark Zuckerberg burning forty five billion dollars 356 00:18:59,760 --> 00:19:02,639 Speaker 1: for no reason. In fact, there's a Yahoo Finance story 357 00:19:02,800 --> 00:19:05,240 Speaker 1: about the fact that all of that money that's allegedly 358 00:19:05,240 --> 00:19:07,800 Speaker 1: going into the metaverse is just organizational loss. It's a 359 00:19:07,880 --> 00:19:12,679 Speaker 1: horribly run place where the management constantly changes. These goddamn people. 360 00:19:13,320 --> 00:19:16,320 Speaker 1: And Amazon, they've never really been in the ideas business. 361 00:19:16,600 --> 00:19:19,240 Speaker 1: They're the founder of a cloud storage cartel with Microsoft, 362 00:19:19,240 --> 00:19:22,640 Speaker 1: Oracle and Google. They work together to set the prices. 363 00:19:22,920 --> 00:19:26,920 Speaker 1: Everyone wins except anyone who isn't them. And big tech 364 00:19:27,000 --> 00:19:29,560 Speaker 1: is lazy because they've all agreed to compete only a 365 00:19:29,640 --> 00:19:31,920 Speaker 1: little bit, never coloring too far out of the lines 366 00:19:31,920 --> 00:19:34,800 Speaker 1: because doing so might expose them to an actual risk, 367 00:19:34,920 --> 00:19:37,480 Speaker 1: like regulation or having to hire people that build new things, 368 00:19:37,560 --> 00:19:42,000 Speaker 1: versus finding increasingly more annoying ways to exploit their current customers. 369 00:19:43,080 --> 00:19:45,560 Speaker 1: And I really am sure that there's a societal shift 370 00:19:45,560 --> 00:19:48,399 Speaker 1: coming against big tech, the natural result of years of 371 00:19:48,520 --> 00:19:51,480 Speaker 1: excess and bullshit of the loudest people in tech trying 372 00:19:51,520 --> 00:19:53,560 Speaker 1: to convince people that a large language model that can 373 00:19:53,560 --> 00:19:57,040 Speaker 1: generate texts and images that kind of suck is basically 374 00:19:57,040 --> 00:20:00,800 Speaker 1: the same thing as an iPhone. I have no idea 375 00:20:00,840 --> 00:20:04,080 Speaker 1: how quickly the shift comes, but it's coming. It's inevitable. 376 00:20:04,920 --> 00:20:08,359 Speaker 1: This lack of big, profitable ideas will eventually lead to 377 00:20:08,400 --> 00:20:11,880 Speaker 1: some sort of reckoning or collapse. In big tech, healthy, 378 00:20:12,000 --> 00:20:14,800 Speaker 1: well run companies, they don't do things like this, let 379 00:20:14,800 --> 00:20:18,120 Speaker 1: alone several different trillion dollar companies all bumbling into each 380 00:20:18,119 --> 00:20:20,760 Speaker 1: other like the Three Stooges, billions of dollars of cash 381 00:20:20,800 --> 00:20:25,280 Speaker 1: falling out of their pockets. It's an absolute disgrace. I 382 00:20:25,320 --> 00:20:28,520 Speaker 1: will say, if the AI bubble really does pop, this 383 00:20:28,560 --> 00:20:31,720 Speaker 1: is a great time to point and laugh. Regular people 384 00:20:31,800 --> 00:20:34,120 Speaker 1: like yourself could see this was bullshit from the beginning. 385 00:20:34,760 --> 00:20:39,119 Speaker 1: Most people find generative AI either briefly interesting or somewhat disgusting, 386 00:20:39,400 --> 00:20:42,000 Speaker 1: and they're capable of seeing that. Stealing the entire Internet 387 00:20:42,040 --> 00:20:43,960 Speaker 1: so you could spend one hundred billion dollars making a 388 00:20:43,960 --> 00:20:47,240 Speaker 1: shitty version of the past is kind of a stupid idea. 389 00:20:47,400 --> 00:20:51,240 Speaker 1: The utility really hasn't been there from the beginning. Describing 390 00:20:51,280 --> 00:20:55,359 Speaker 1: what chat GBT is is actually kind of difficult, and 391 00:20:55,480 --> 00:20:58,119 Speaker 1: any technology that's going to be worth trillions of dollars 392 00:20:58,560 --> 00:21:00,959 Speaker 1: should not have the user saying, wait, what does this 393 00:21:01,040 --> 00:21:04,240 Speaker 1: do exactly? And I really feel like regular people notice this. 394 00:21:04,520 --> 00:21:07,080 Speaker 1: If this was as magical as it was, regular people 395 00:21:07,119 --> 00:21:09,240 Speaker 1: would be lining up to use it. And while it 396 00:21:09,280 --> 00:21:11,960 Speaker 1: has a lot of users. I don't know anyone who's 397 00:21:12,040 --> 00:21:17,159 Speaker 1: fanatical over it other than AI fanatics. It's just very sad. 398 00:21:18,720 --> 00:21:20,919 Speaker 1: But regular people aren't the only ones. And I have 399 00:21:21,000 --> 00:21:24,000 Speaker 1: criticized journalists for missing this, but I want to call 400 00:21:24,080 --> 00:21:26,920 Speaker 1: some people out in a good way. Economists, analysts and 401 00:21:27,000 --> 00:21:29,520 Speaker 1: journalists have been saying this isn't the future, like I said. 402 00:21:29,720 --> 00:21:32,040 Speaker 1: Shira a v Day in March twenty twenty three, mere 403 00:21:32,080 --> 00:21:35,640 Speaker 1: months after chat GPT's launch, said it wasn't magic. For example, 404 00:21:35,640 --> 00:21:39,320 Speaker 1: in the Washington Post, Shearer the Goat, Alex Kantrowitz and 405 00:21:39,320 --> 00:21:42,080 Speaker 1: Douglas Gorman reported in August fifteenth to twenty twenty three 406 00:21:42,080 --> 00:21:44,240 Speaker 1: that chat gpt isn't good enough to take jobs and 407 00:21:44,280 --> 00:21:46,800 Speaker 1: it's unlikely to cause mass layoffs. The hot takes have 408 00:21:46,880 --> 00:21:49,840 Speaker 1: run into reality a little bit early, but you're correct. 409 00:21:50,560 --> 00:21:53,840 Speaker 1: Paris Mark's called the chat GPT revolution another take fantasy 410 00:21:53,840 --> 00:21:56,240 Speaker 1: in July twenty twenty three, and Gary Marcus worried about 411 00:21:56,240 --> 00:21:58,639 Speaker 1: Generative AI being a dud in August twenty twenty three. 412 00:21:58,960 --> 00:22:01,800 Speaker 1: While this is by no means an exhaustive list, it's 413 00:22:01,840 --> 00:22:05,320 Speaker 1: important to note that people have been saying this. People 414 00:22:05,359 --> 00:22:10,120 Speaker 1: with big platforms and the markets chose to ignore them 415 00:22:10,640 --> 00:22:13,919 Speaker 1: as an aside. A year later, when many outlets were 416 00:22:13,920 --> 00:22:16,280 Speaker 1: still tripping over themselves to say that generative AI was 417 00:22:16,280 --> 00:22:19,040 Speaker 1: the future. Sure A Viday of The Washington Post asked, 418 00:22:19,119 --> 00:22:21,320 Speaker 1: if these chatbots are supposed to be magical, why are 419 00:22:21,359 --> 00:22:23,680 Speaker 1: so many of them as dumb as rocks. It's important 420 00:22:23,680 --> 00:22:25,800 Speaker 1: to call out when a national newspaper actually does the 421 00:22:25,840 --> 00:22:27,760 Speaker 1: work that needs to be done to critique the powerful 422 00:22:28,000 --> 00:22:32,120 Speaker 1: good works here, Open invites come on the show. At 423 00:22:32,119 --> 00:22:34,520 Speaker 1: this point, it seems inevitable that the bubble will burst, 424 00:22:34,600 --> 00:22:37,560 Speaker 1: that the acceleration has stopped, that the funding won't be there, 425 00:22:37,560 --> 00:22:39,359 Speaker 1: and that the markets taste for the billions of dollars 426 00:22:39,359 --> 00:22:42,320 Speaker 1: it's going to take to make generative AI. Some indeterminate 427 00:22:42,400 --> 00:22:47,080 Speaker 1: level of better just really isn't there. By all means 428 00:22:47,119 --> 00:22:49,199 Speaker 1: tell me I'm wrong. You certainly do on Twitter, and 429 00:22:49,240 --> 00:22:52,320 Speaker 1: you certainly do on email. And I'd be genuinely intrigued 430 00:22:52,320 --> 00:22:55,359 Speaker 1: to see how open ai wriggles out of this jam, 431 00:22:55,560 --> 00:22:58,000 Speaker 1: especially because it would mean they made a meaningful product 432 00:22:58,040 --> 00:22:59,720 Speaker 1: that makes money when they don't have the chips to 433 00:22:59,760 --> 00:23:02,880 Speaker 1: train GPT five, a model that likely still wouldn't change 434 00:23:02,880 --> 00:23:05,439 Speaker 1: the game enough to matter. I mean, what if they 435 00:23:05,480 --> 00:23:07,720 Speaker 1: have a new architecture, I guess, but I refuse to 436 00:23:07,720 --> 00:23:10,320 Speaker 1: give them credit for something they haven't done yet. And 437 00:23:10,359 --> 00:23:12,600 Speaker 1: in the event that Open Ai or Anthropic is able 438 00:23:12,600 --> 00:23:15,399 Speaker 1: to raise again, we'll see the cycle continue another quarter. 439 00:23:15,520 --> 00:23:18,760 Speaker 1: I'm confident, at which point something crazy would need to 440 00:23:18,760 --> 00:23:21,200 Speaker 1: happen to make any of this worth it. If these 441 00:23:21,240 --> 00:23:24,439 Speaker 1: companies can actually raise, they'll keep the bubble inflated. But 442 00:23:24,560 --> 00:23:28,280 Speaker 1: it really is borrowed time. There's enough money to fund 443 00:23:28,359 --> 00:23:31,840 Speaker 1: them both maybe one more time, and after that, where 444 00:23:31,880 --> 00:23:35,280 Speaker 1: is the profit? Where's this going? And it's really worth 445 00:23:35,280 --> 00:23:37,800 Speaker 1: considering that the narrative is bad and that there is 446 00:23:37,960 --> 00:23:40,399 Speaker 1: very little that anyone can really do to change it 447 00:23:40,440 --> 00:23:44,720 Speaker 1: outside of something truly incredible. These companies can play for time. 448 00:23:44,840 --> 00:23:47,160 Speaker 1: Hope that satch In a Dele and sun A Peshai 449 00:23:47,200 --> 00:23:49,639 Speaker 1: stan pats say that they have great faith in the 450 00:23:49,680 --> 00:23:52,840 Speaker 1: promise of AI, but I really don't think they get 451 00:23:52,880 --> 00:23:56,359 Speaker 1: another series of earnings for big tech like this. In 452 00:23:56,400 --> 00:24:00,679 Speaker 1: many ways, though, this is sam Ortman's greatest gambit, and 453 00:24:00,760 --> 00:24:03,200 Speaker 1: so much lies and Open Eyes back as a result. 454 00:24:03,920 --> 00:24:06,399 Speaker 1: This is the moment we find out if Sam Altman 455 00:24:06,480 --> 00:24:09,840 Speaker 1: has really got it. Whether it's worth agonizing and wringing 456 00:24:09,840 --> 00:24:12,800 Speaker 1: our hands about how scary he is, or whether he's 457 00:24:12,840 --> 00:24:15,360 Speaker 1: just able to hoodwink enough credulous billionaires that have their 458 00:24:15,359 --> 00:24:19,800 Speaker 1: hands well did to eternal money printers. Impress me, Samuel, 459 00:24:20,400 --> 00:24:25,040 Speaker 1: Wake up, Sammy, mister alt men, bring me the technology. 460 00:24:25,160 --> 00:24:28,040 Speaker 1: Show me what you've got, Show me your great works, 461 00:24:28,280 --> 00:24:30,160 Speaker 1: Show me the works you have buried in that non 462 00:24:30,280 --> 00:24:34,080 Speaker 1: profit for profit monstrosity. You've used the trick other visionless 463 00:24:34,119 --> 00:24:38,000 Speaker 1: power brokers into giving you their entire futures for a 464 00:24:38,040 --> 00:24:41,040 Speaker 1: dead end friend best known for an absenteism that raankled 465 00:24:41,040 --> 00:24:43,359 Speaker 1: his peers and some of the startups he was supposed 466 00:24:43,400 --> 00:24:47,240 Speaker 1: to nurture. Show me what you've got, Sam, show us all. 467 00:24:47,800 --> 00:24:50,719 Speaker 1: You're so smart. You allegedly can solve problems in a 468 00:24:50,760 --> 00:24:54,879 Speaker 1: single phone call. Everyone talks about your big brain. What's 469 00:24:54,960 --> 00:24:58,879 Speaker 1: the big play, big guy? You've got maybe six months 470 00:24:58,920 --> 00:25:02,159 Speaker 1: and that's the most general estimate I've got. What is it? 471 00:25:02,680 --> 00:25:08,320 Speaker 1: What's the plan? Wait? What's that? Search? GPT? Search? GPT? Sam? 472 00:25:08,359 --> 00:25:11,520 Speaker 1: You're crazy, You're crazy, mister ortman. There's no way you 473 00:25:11,560 --> 00:25:15,000 Speaker 1: can build a Google Search competitor. Trust me. I know, no, really, 474 00:25:15,000 --> 00:25:17,320 Speaker 1: I worked it out. I spent a few hours looking 475 00:25:17,400 --> 00:25:19,639 Speaker 1: up Google's financial reports over the last decade or so, 476 00:25:19,720 --> 00:25:21,600 Speaker 1: trying to figure out how much the company spends and 477 00:25:21,640 --> 00:25:25,320 Speaker 1: operating costs and capital expenditures just related to search alone. 478 00:25:25,840 --> 00:25:28,200 Speaker 1: Due to the way Google structures it's financials, I had 479 00:25:28,200 --> 00:25:31,040 Speaker 1: to make some guesses, some more informed than others. But 480 00:25:31,119 --> 00:25:33,840 Speaker 1: I reckon it costs Google about thirty billion dollars a 481 00:25:33,920 --> 00:25:36,440 Speaker 1: year to run their search product, and that's not taken 482 00:25:36,480 --> 00:25:39,440 Speaker 1: into account the cost of the generative AI powered search summaries. 483 00:25:40,000 --> 00:25:42,400 Speaker 1: All right, Sam, here we are. So that's thirty billion 484 00:25:42,440 --> 00:25:44,560 Speaker 1: dollars a year to create something that can scale to 485 00:25:44,640 --> 00:25:46,879 Speaker 1: the size of Google Search. And that's just the start 486 00:25:46,880 --> 00:25:50,040 Speaker 1: you with me, Man, things do not get easier. You see, 487 00:25:50,080 --> 00:25:52,760 Speaker 1: you're not just building a Google Search competitor with search 488 00:25:52,800 --> 00:25:55,040 Speaker 1: gpt as. You also have to build the ad take 489 00:25:55,080 --> 00:25:57,359 Speaker 1: infrastructure to actually make money off of it, which you 490 00:25:57,400 --> 00:26:00,000 Speaker 1: are not good at. Okay, but you've got to copy 491 00:26:00,080 --> 00:26:02,600 Speaker 1: features from companies like double Click, which Google acquired in 492 00:26:02,600 --> 00:26:05,720 Speaker 1: two thousand and seven for three point one billion dollars 493 00:26:05,720 --> 00:26:08,760 Speaker 1: and pretty much underpins the whole operation. Or maybe you 494 00:26:08,760 --> 00:26:10,960 Speaker 1: have to build something like AdWords, which Google got when 495 00:26:10,960 --> 00:26:14,200 Speaker 1: it acquired Applied Semantics two thousand and three. You will 496 00:26:14,240 --> 00:26:16,720 Speaker 1: need something like both of these, and that's going to 497 00:26:16,800 --> 00:26:20,000 Speaker 1: cost you, in the best light, in the best day ever, 498 00:26:20,200 --> 00:26:22,320 Speaker 1: about three billion, or you're gonna have to outsource it, 499 00:26:22,359 --> 00:26:25,120 Speaker 1: which will make it worse and less profitable because outsourced 500 00:26:25,119 --> 00:26:26,760 Speaker 1: companies are going to take a cut of your revenue, 501 00:26:26,800 --> 00:26:30,480 Speaker 1: thus reducing your margins, which are already really bad. Then 502 00:26:30,600 --> 00:26:33,399 Speaker 1: you've got to hire experienced ad personnel and use that 503 00:26:33,440 --> 00:26:36,240 Speaker 1: ad tech infrastructure you built. You built that right to 504 00:26:36,400 --> 00:26:39,080 Speaker 1: enter a market. Google will use their entire cartel power 505 00:26:39,080 --> 00:26:43,119 Speaker 1: to kick you out of Wait what was that? Sam? Wait? What? Right? Right? 506 00:26:43,200 --> 00:26:45,280 Speaker 1: You've got to make money, and ideally more of it 507 00:26:45,320 --> 00:26:47,600 Speaker 1: than you spend something. As I've mentioned, you have some 508 00:26:47,760 --> 00:26:51,360 Speaker 1: issues with. Like I said, Google Search costs an absolute 509 00:26:51,359 --> 00:26:53,680 Speaker 1: shit toun to run thirty billion dollars easy, and that's 510 00:26:53,720 --> 00:26:57,200 Speaker 1: with decades of optimization, a coding language that they invented, 511 00:26:57,320 --> 00:27:00,600 Speaker 1: thousands of miles of underground cables, and decades experience in 512 00:27:00,680 --> 00:27:02,719 Speaker 1: doing a thing you're just learning to do, and they 513 00:27:02,760 --> 00:27:05,520 Speaker 1: have PhDs who specialize it. Wh've been at the company decades. 514 00:27:05,920 --> 00:27:09,040 Speaker 1: You did all that? Whoa? Okay? Sam? I'm impressed. You're 515 00:27:09,040 --> 00:27:11,119 Speaker 1: on a roll. You can do this. I believe in you, 516 00:27:11,160 --> 00:27:14,520 Speaker 1: mister Altman. Now just checking, how many AD salespeople do 517 00:27:14,560 --> 00:27:18,040 Speaker 1: you have? What's that? None? You don't have any ad 518 00:27:18,080 --> 00:27:20,800 Speaker 1: salespeople for your search engine, Sam, You're gonna need them. 519 00:27:21,119 --> 00:27:24,000 Speaker 1: Google had over thirty thousand AD salespeople before. It's the 520 00:27:24,040 --> 00:27:26,280 Speaker 1: last round of layoffs. You're not gonna make this thing 521 00:27:26,320 --> 00:27:29,120 Speaker 1: profitable without ads. In fact, I don't think Sam, you're 522 00:27:29,119 --> 00:27:32,240 Speaker 1: gonna make this search product profitable at all. But you 523 00:27:32,320 --> 00:27:34,960 Speaker 1: know what may or may not be profitable. One of 524 00:27:35,000 --> 00:27:37,640 Speaker 1: our advertisers who will soon invade your living room, car, 525 00:27:37,720 --> 00:27:40,200 Speaker 1: or other place you listen to podcasts filling your various 526 00:27:40,200 --> 00:27:57,120 Speaker 1: crevices with commerce. Behold, they're bounty. All right, right, we're back, 527 00:27:57,160 --> 00:27:59,720 Speaker 1: and I'll dispense with the sam Ormond ben so one 528 00:27:59,880 --> 00:28:02,480 Speaker 1: was looking up Google's financials, As I mentioned, I was 529 00:28:02,520 --> 00:28:04,560 Speaker 1: able to make a few educated guesses, and I've got 530 00:28:04,560 --> 00:28:06,840 Speaker 1: a hunch that Google has about a thirty five percent 531 00:28:06,880 --> 00:28:09,400 Speaker 1: profit margin, which is pretty good. But that's with over 532 00:28:09,560 --> 00:28:12,320 Speaker 1: two decades of infrastructure and a near monopoly on the 533 00:28:12,320 --> 00:28:16,560 Speaker 1: search market. Open Ai has none of those things. They 534 00:28:16,560 --> 00:28:18,879 Speaker 1: don't really own their infrastructure at all, as I mentioned, 535 00:28:19,440 --> 00:28:22,439 Speaker 1: and it's going to be relying entirely on GPT for 536 00:28:22,480 --> 00:28:25,240 Speaker 1: the most part, which is insanely power and compute intensive 537 00:28:25,640 --> 00:28:27,280 Speaker 1: and thus cost a lot more. And as I've said 538 00:28:27,320 --> 00:28:30,840 Speaker 1: a billion times already, GPT and other transformer based architecture 539 00:28:30,880 --> 00:28:34,199 Speaker 1: is deeply unprofitable, unsustainable. And that's before you even get 540 00:28:34,240 --> 00:28:37,680 Speaker 1: to the hallucinations, the hallucinations that underpin your search engine. 541 00:28:37,720 --> 00:28:40,760 Speaker 1: You need to be profitable, which isn't profitable. Oh boy, 542 00:28:40,880 --> 00:28:43,760 Speaker 1: this is really bad for them. Assuming the bulk of 543 00:28:43,800 --> 00:28:45,960 Speaker 1: the costs associated with running Google Search comes from the 544 00:28:46,040 --> 00:28:49,000 Speaker 1: vast number of service required to actually power it, it's 545 00:28:49,040 --> 00:28:52,080 Speaker 1: hard to imagine how anything open ai does could even 546 00:28:52,640 --> 00:28:55,520 Speaker 1: ever reach a profit, even if they had the monopoly 547 00:28:55,560 --> 00:28:59,400 Speaker 1: Google has. Now, I'm confident that open ai is trying 548 00:28:59,440 --> 00:29:01,560 Speaker 1: to raise cap or right now on the idea that 549 00:29:01,600 --> 00:29:04,320 Speaker 1: search GPT is able to create a meaningful competitor to 550 00:29:04,400 --> 00:29:07,640 Speaker 1: the most profitable software business ever made, one that Google 551 00:29:07,680 --> 00:29:11,040 Speaker 1: will likely protect with various monopolistic weapons like its treacher 552 00:29:11,160 --> 00:29:14,440 Speaker 1: trove of patents. While the Department of Justice did just 553 00:29:14,480 --> 00:29:16,760 Speaker 1: declare that Google had a monopoly over the search market, 554 00:29:16,800 --> 00:29:19,240 Speaker 1: and I'll get to that in an upcoming episode. It 555 00:29:19,280 --> 00:29:21,240 Speaker 1: will take years for things to move on that front, 556 00:29:21,400 --> 00:29:25,280 Speaker 1: and nothing about this situation fixes any of open AI's problems. 557 00:29:26,640 --> 00:29:29,200 Speaker 1: On top of that, Google spends, as I've previously mentioned, 558 00:29:29,200 --> 00:29:31,880 Speaker 1: billions of dollars on hardware and property. An open ai 559 00:29:31,960 --> 00:29:34,560 Speaker 1: would likely have to build infrastructure that dwarfs even the 560 00:29:34,600 --> 00:29:37,520 Speaker 1: outlays of Generative AI to scale a search engine that 561 00:29:37,520 --> 00:29:40,320 Speaker 1: could handle even hundreds of millions that loan the billions 562 00:29:40,360 --> 00:29:43,560 Speaker 1: of people that use Google Search. It could get Microsoft 563 00:29:43,600 --> 00:29:45,560 Speaker 1: to build it or lease it from the likes of Oracle, 564 00:29:45,880 --> 00:29:49,120 Speaker 1: and Microsoft could probably pay for it too, but that 565 00:29:49,160 --> 00:29:51,880 Speaker 1: would be another massive increase in capital expenditures at a 566 00:29:51,880 --> 00:29:54,200 Speaker 1: time when the markets are waiting for the stuff to 567 00:29:54,280 --> 00:29:57,080 Speaker 1: make money, rather than the companies to spend more money 568 00:29:57,080 --> 00:29:59,920 Speaker 1: on it. And I also should note that Microsoft reason 569 00:30:00,120 --> 00:30:02,760 Speaker 1: he described open ai as a competitor in the search business, 570 00:30:02,960 --> 00:30:05,200 Speaker 1: and it's unlikely to kill bing, which has been profitable 571 00:30:05,240 --> 00:30:08,840 Speaker 1: since twenty sixteen. Let's be honest, though this is all 572 00:30:08,880 --> 00:30:11,880 Speaker 1: fan fection. Open ai isn't competing with Google Search. It 573 00:30:11,920 --> 00:30:14,520 Speaker 1: would cost too much money, and the underlying technology is 574 00:30:14,520 --> 00:30:17,160 Speaker 1: neither good enough nor efficient enough to actually support a 575 00:30:17,160 --> 00:30:20,840 Speaker 1: search engine at scale, and even if it was, is 576 00:30:20,880 --> 00:30:24,800 Speaker 1: it better? Probably not. What about chat? GPT says that 577 00:30:24,800 --> 00:30:27,920 Speaker 1: it's going to be more reliable than Google Search, and God, 578 00:30:28,080 --> 00:30:31,320 Speaker 1: all of this is such a boring fucking face. We're 579 00:30:31,320 --> 00:30:35,000 Speaker 1: not watching startups mess around Silicon Valley experimenting in ways 580 00:30:35,000 --> 00:30:38,239 Speaker 1: to fix weird problems. We're watching big tech outsource their 581 00:30:38,240 --> 00:30:40,520 Speaker 1: research and development costs to one of a few hand 582 00:30:40,560 --> 00:30:43,160 Speaker 1: selected con artists that they'll drop the second that Wall 583 00:30:43,200 --> 00:30:47,680 Speaker 1: Street gets pissy. Why even pretend anymore? Anthropic is propped 584 00:30:47,720 --> 00:30:49,960 Speaker 1: up by Google and Amazon, who invested billions of dollars 585 00:30:50,000 --> 00:30:52,760 Speaker 1: in cloud credits. Microsoft props up open Ai to the 586 00:30:52,800 --> 00:30:56,000 Speaker 1: point that the Nvidia chips crisis affected here because Daddy 587 00:30:56,000 --> 00:30:59,800 Speaker 1: Microsoft was buying them for open Ai to use, and Google, Amazon, 588 00:30:59,840 --> 00:31:02,440 Speaker 1: and Microsoft are all making their own models, to the 589 00:31:02,440 --> 00:31:05,440 Speaker 1: point that Microsoft has already reported, like I said, that 590 00:31:05,480 --> 00:31:09,640 Speaker 1: they see open ai as a competitor. It's just it's 591 00:31:09,680 --> 00:31:13,560 Speaker 1: just also craven. Big tech funded a cloud dependent product 592 00:31:13,600 --> 00:31:15,680 Speaker 1: that costs so much more than anything else in the 593 00:31:15,760 --> 00:31:19,280 Speaker 1: market based on a spacious hype about what artificial intelligence 594 00:31:19,320 --> 00:31:22,320 Speaker 1: could or could not do. They do so as a 595 00:31:22,360 --> 00:31:25,080 Speaker 1: means of selling a new product, one that, while not 596 00:31:25,160 --> 00:31:27,760 Speaker 1: particularly useful, also means that they can charge a lot 597 00:31:27,760 --> 00:31:30,680 Speaker 1: for the cloud compute service it runs on, creating demand 598 00:31:30,760 --> 00:31:33,080 Speaker 1: for a product that pays them money just for existing, 599 00:31:33,320 --> 00:31:36,600 Speaker 1: allowing them to vastly expand their physical data center locations, 600 00:31:36,920 --> 00:31:41,880 Speaker 1: further enforcing their monopolies over the cloud storage industry. Microsoft, Google, Meta, 601 00:31:41,960 --> 00:31:44,160 Speaker 1: and Amazon created a new way to turn money into 602 00:31:44,200 --> 00:31:46,560 Speaker 1: more money investing in a tech that required you to 603 00:31:46,600 --> 00:31:49,560 Speaker 1: pay them all a lot of money for cloud computing 604 00:31:49,600 --> 00:31:52,520 Speaker 1: services to run on. Like I said, the problem, it seems, 605 00:31:52,640 --> 00:31:54,640 Speaker 1: was that none of them were willing to consider a 606 00:31:54,680 --> 00:31:57,600 Speaker 1: world where this stuff never really turned into anything remarkable 607 00:31:57,640 --> 00:32:00,719 Speaker 1: other than in its waste. And a lack of creativity 608 00:32:00,720 --> 00:32:02,800 Speaker 1: in the tallest towers of the valley in Redmond has 609 00:32:02,800 --> 00:32:04,840 Speaker 1: allowed the richest men in the industry to be conned 610 00:32:04,840 --> 00:32:07,320 Speaker 1: by a snake oil machine that loses billions of billions 611 00:32:07,320 --> 00:32:10,720 Speaker 1: of dollars. To be clear, there's nothing wrong on its 612 00:32:10,720 --> 00:32:14,560 Speaker 1: face inherently with generative AI. It does interesting things with documents, 613 00:32:14,560 --> 00:32:16,880 Speaker 1: and I fully believe there'll be some low five version 614 00:32:17,080 --> 00:32:20,480 Speaker 1: of generative AI that plugs around silently once the hyperscalers 615 00:32:20,520 --> 00:32:24,440 Speaker 1: piss off. My problem is that it's so horribly wasteful 616 00:32:24,680 --> 00:32:27,280 Speaker 1: to put all of this money into these large language models, 617 00:32:27,480 --> 00:32:29,880 Speaker 1: the ones that steal everything, that require so much energy 618 00:32:29,920 --> 00:32:33,200 Speaker 1: and create so little in return. Now there reason not 619 00:32:33,280 --> 00:32:36,800 Speaker 1: brought up previously. The whole calculation of open AI needing 620 00:32:36,840 --> 00:32:39,360 Speaker 1: five to ten billion dollars to survive is because I 621 00:32:39,360 --> 00:32:43,040 Speaker 1: wanted to make it clear how expensive them staying in places. 622 00:32:43,720 --> 00:32:46,640 Speaker 1: That is the money they will need to continue building 623 00:32:46,640 --> 00:32:49,880 Speaker 1: the thing they are doing today. They don't have anything 624 00:32:50,280 --> 00:32:54,000 Speaker 1: in reserve. I am confident about it. Maybe they can survive, 625 00:32:54,480 --> 00:32:57,440 Speaker 1: Maybe they can surprise me. Maybe that can happen, But 626 00:32:57,480 --> 00:33:01,320 Speaker 1: I just don't really see it happening. Greg Brockman's gone, 627 00:33:01,320 --> 00:33:03,160 Speaker 1: and I mean, he was just another non technical guy. 628 00:33:03,280 --> 00:33:05,080 Speaker 1: What's going to happen. You think Mirror Marat's is going 629 00:33:05,160 --> 00:33:07,040 Speaker 1: to come up with something, You think Sura is going 630 00:33:07,120 --> 00:33:09,880 Speaker 1: to do it. No, they don't have anything. And I 631 00:33:09,920 --> 00:33:13,600 Speaker 1: realize it's so strange to say this, especially if you're 632 00:33:13,560 --> 00:33:15,560 Speaker 1: a member of the media listening to this. I know 633 00:33:15,600 --> 00:33:18,400 Speaker 1: it's strange to accept this. But what if they're just 634 00:33:18,440 --> 00:33:21,440 Speaker 1: full of shit? What if this is the best they've got? 635 00:33:22,240 --> 00:33:25,120 Speaker 1: How do we deal with this? And I think we 636 00:33:25,280 --> 00:33:28,360 Speaker 1: start by accepting that all of this was completely and 637 00:33:28,480 --> 00:33:33,360 Speaker 1: utterly ridiculous. We live in incredible times. We're at a 638 00:33:33,360 --> 00:33:36,520 Speaker 1: boiling point in the tech industry's history where these sins 639 00:33:36,520 --> 00:33:39,040 Speaker 1: of excess and this total lack of vision of crippled 640 00:33:39,160 --> 00:33:42,440 Speaker 1: big tis ability to make new things. Things have been 641 00:33:42,480 --> 00:33:45,840 Speaker 1: too easy, too comfortable, and too frictionless for too long 642 00:33:46,000 --> 00:33:49,160 Speaker 1: for people like Sundhar Peshai and Sachinidella, and even people 643 00:33:49,200 --> 00:33:51,320 Speaker 1: like Larry Page and Sergey Brin, who are the only 644 00:33:51,360 --> 00:33:54,960 Speaker 1: two people who can fire Sandar from Google. Tech needs 645 00:33:55,000 --> 00:33:57,480 Speaker 1: to go back to creating interesting and cool things and 646 00:33:57,520 --> 00:34:00,000 Speaker 1: having companies run by people that actually care about build 647 00:34:00,000 --> 00:34:03,560 Speaker 1: holding them. Technology is capable of solving real problems, or 648 00:34:03,560 --> 00:34:05,720 Speaker 1: at least it has been in the past and proven 649 00:34:05,800 --> 00:34:07,800 Speaker 1: time and time again how profitable it can be to 650 00:34:07,840 --> 00:34:10,480 Speaker 1: do so, and I believe it can do just fine 651 00:34:10,560 --> 00:34:13,080 Speaker 1: without having four or five companies that do every single 652 00:34:13,120 --> 00:34:15,440 Speaker 1: thing that technology can do at the same time in 653 00:34:15,520 --> 00:34:19,160 Speaker 1: hopes that they can monopolize it. I don't think the 654 00:34:19,200 --> 00:34:22,719 Speaker 1: most powerful people in tech realize how frustrated people are, 655 00:34:23,080 --> 00:34:26,680 Speaker 1: how deeply affected by technology everyone is, even those who 656 00:34:26,760 --> 00:34:31,040 Speaker 1: don't identify as tech nerds or enthusiasts, and I think 657 00:34:31,120 --> 00:34:34,160 Speaker 1: men like Sandhar Peshai and Sachi Nadella are deeply unaware 658 00:34:34,160 --> 00:34:36,960 Speaker 1: of how ready people are for a change, how disgusted 659 00:34:36,960 --> 00:34:38,880 Speaker 1: they are by the crater in quality of the average 660 00:34:38,920 --> 00:34:41,720 Speaker 1: tech product at a time when executives have never been richer. 661 00:34:42,840 --> 00:34:46,359 Speaker 1: Regular consumers know that Google Search sucks. They know that 662 00:34:46,400 --> 00:34:49,480 Speaker 1: Facebook sucks. They know that websites are worse, that they 663 00:34:49,600 --> 00:34:51,839 Speaker 1: visit certain ones and there's eighteen pop ups and your 664 00:34:51,840 --> 00:34:55,240 Speaker 1: phone's really hot. They know that their iPhones are randomly 665 00:34:55,239 --> 00:34:57,919 Speaker 1: trying to make them buy Apple News, and they know 666 00:34:58,239 --> 00:35:01,680 Speaker 1: that there are people responsible. Tim Cook, EDDIQA Apple By 667 00:35:01,719 --> 00:35:04,520 Speaker 1: the way, welcome to come on the show, But guys, 668 00:35:04,520 --> 00:35:06,879 Speaker 1: stop trying to sell me Apple News. I don't want it. 669 00:35:07,040 --> 00:35:09,319 Speaker 1: I don't want any of this Apple shit. Don't make 670 00:35:09,360 --> 00:35:11,600 Speaker 1: me use it unless I have to other than I message. 671 00:35:12,000 --> 00:35:15,239 Speaker 1: But moving on, there are real targets here. There are 672 00:35:15,239 --> 00:35:18,640 Speaker 1: people that need to go to fix things. S Darpishai 673 00:35:18,719 --> 00:35:23,880 Speaker 1: and Sachinidella must go. They are no longer trustworthy stewards 674 00:35:23,920 --> 00:35:26,319 Speaker 1: of the future and have inspired hundreds of billions of 675 00:35:26,360 --> 00:35:29,879 Speaker 1: dollars of waste in the hearts of their equally myopic competitors. 676 00:35:30,440 --> 00:35:33,120 Speaker 1: These men do not care about their customers, they don't 677 00:35:33,280 --> 00:35:40,920 Speaker 1: fear competition, and the government should break every single hyperscala up, Google, Meta, Amazon, Apple, Oracle, 678 00:35:41,120 --> 00:35:43,160 Speaker 1: all of them, and find ways to stop them ever 679 00:35:43,200 --> 00:35:46,960 Speaker 1: growing so large. In the future. Maybe we shouldn't have 680 00:35:47,280 --> 00:35:51,880 Speaker 1: such large tech companies. Maybe we don't need multiple trillion 681 00:35:51,920 --> 00:35:55,920 Speaker 1: dollar juggernauts that do everything in a shitty way. Maybe 682 00:35:55,960 --> 00:35:59,880 Speaker 1: we don't need companies that do online advertising and or 683 00:36:00,080 --> 00:36:02,680 Speaker 1: so provide a platform, because that kind of monopoly leads 684 00:36:02,680 --> 00:36:05,759 Speaker 1: to insane incentives like you're seeing with Google Search and Facebook, 685 00:36:06,360 --> 00:36:09,080 Speaker 1: and we all deserve better, and the way to get 686 00:36:09,080 --> 00:36:11,239 Speaker 1: there is to say the names of those responsible, to 687 00:36:11,320 --> 00:36:14,880 Speaker 1: document their disgraceful acts, and call for accountability again and 688 00:36:14,920 --> 00:36:18,440 Speaker 1: again and again and again until something actually happens and changes. 689 00:36:19,400 --> 00:36:22,719 Speaker 1: And it is changing. The narrative is changing, and soon 690 00:36:22,760 --> 00:36:25,959 Speaker 1: it will shift more aggressively and more violently against big tech. 691 00:36:26,040 --> 00:36:27,320 Speaker 1: And I think this is a great thing for the 692 00:36:27,320 --> 00:36:30,879 Speaker 1: world and ultimately to take industry at large. Now as 693 00:36:30,880 --> 00:36:35,279 Speaker 1: we see the bubble, the AI bubble collapse, and must 694 00:36:35,320 --> 00:36:37,200 Speaker 1: be clear, none of this is a victory lap for me. 695 00:36:37,280 --> 00:36:40,239 Speaker 1: This is not actually, how I wish this went. I 696 00:36:40,280 --> 00:36:42,399 Speaker 1: wish this bubble had burst earlier. It would have caused 697 00:36:42,480 --> 00:36:45,120 Speaker 1: less harm to the environment, to the markets, to the 698 00:36:45,160 --> 00:36:47,840 Speaker 1: freelancers who lose their jobs, to the artists who email 699 00:36:47,920 --> 00:36:51,200 Speaker 1: me every week. It would have wasted less money, taken 700 00:36:51,280 --> 00:36:54,120 Speaker 1: less attention away from the rest of the industry, and 701 00:36:54,200 --> 00:36:57,960 Speaker 1: allowed Sam Wortman to accumulate less money and power. Instead, 702 00:36:58,000 --> 00:37:00,279 Speaker 1: this bubble has been allowed to swell so long it 703 00:37:00,320 --> 00:37:03,200 Speaker 1: will only emphasize how bereft of ideas big tech has become, 704 00:37:03,360 --> 00:37:06,400 Speaker 1: and the consequences are a growing distent against tech from 705 00:37:06,440 --> 00:37:09,760 Speaker 1: the general public. And this era was ultimately the result 706 00:37:09,760 --> 00:37:11,439 Speaker 1: of a lack of fear in the hearts of big 707 00:37:11,480 --> 00:37:13,759 Speaker 1: tech and a deeply held belief that their positions were 708 00:37:13,760 --> 00:37:17,520 Speaker 1: permanent and their customers were trapped. My purpose in documenting 709 00:37:17,520 --> 00:37:19,960 Speaker 1: this in what feels at times like narrating the end 710 00:37:19,960 --> 00:37:22,120 Speaker 1: of the world. I'm not looking to be a pundit 711 00:37:22,280 --> 00:37:23,719 Speaker 1: or a talk in head, or a person who wants 712 00:37:23,760 --> 00:37:25,960 Speaker 1: to be given credit for calling out the bubble, as 713 00:37:26,000 --> 00:37:28,840 Speaker 1: none of these serve you, the customer, the listener, or 714 00:37:28,920 --> 00:37:32,680 Speaker 1: history itself. What I say may be imperfect, but it 715 00:37:32,719 --> 00:37:34,759 Speaker 1: exists to try and give you a clear understanding of 716 00:37:34,760 --> 00:37:38,040 Speaker 1: what's happening. And yes, it's a little emotionally charged and 717 00:37:38,160 --> 00:37:40,960 Speaker 1: very clear in its biases. I believe that the tech 718 00:37:41,000 --> 00:37:43,719 Speaker 1: industry has lost its love and curiosity, and I find 719 00:37:43,719 --> 00:37:46,480 Speaker 1: it disgusting, and the best way to fight back is 720 00:37:46,520 --> 00:37:49,759 Speaker 1: to studiously explain things as they happen from my perspective 721 00:37:49,840 --> 00:37:52,880 Speaker 1: at length, but with I hope a clarity that doesn't 722 00:37:52,880 --> 00:37:57,160 Speaker 1: require a ton of technical knowledge. As things have accelerated, 723 00:37:57,320 --> 00:37:59,560 Speaker 1: I've become a little more introspective about the role of 724 00:37:59,600 --> 00:38:03,319 Speaker 1: my podcas as, my newsletter and my posts. I don't 725 00:38:03,360 --> 00:38:06,000 Speaker 1: have a particularly profound conclusion other than that I believe 726 00:38:06,040 --> 00:38:07,919 Speaker 1: that I have a brevity and clarity about the tech 727 00:38:08,000 --> 00:38:10,960 Speaker 1: industry that many don't. And I also find that I'm 728 00:38:11,000 --> 00:38:13,520 Speaker 1: extremely fast at writing. And I also find this all 729 00:38:13,640 --> 00:38:16,440 Speaker 1: very interesting because I'm a horrible little goblin, and I 730 00:38:16,440 --> 00:38:18,799 Speaker 1: don't talk for nearly two hours or write thousands of 731 00:38:18,800 --> 00:38:21,480 Speaker 1: words because I desperately need the attention, which I do love. 732 00:38:22,280 --> 00:38:25,280 Speaker 1: I do it for the thrill I like covering this stuff, 733 00:38:25,560 --> 00:38:27,799 Speaker 1: and because if I don't say these things out loud, 734 00:38:27,840 --> 00:38:29,960 Speaker 1: the cats in my brain will eventually try and emerge. 735 00:38:30,560 --> 00:38:33,040 Speaker 1: But seriously, though I do these things as I need to. 736 00:38:33,400 --> 00:38:36,319 Speaker 1: This is important to me. Maybe I'm fighting back. Maybe 737 00:38:36,360 --> 00:38:38,799 Speaker 1: I just like creating. I don't really know. This is 738 00:38:38,880 --> 00:38:42,359 Speaker 1: me processing watching an industry I care deeply about get 739 00:38:42,440 --> 00:38:45,000 Speaker 1: ransecked again and again by people who don't seem to 740 00:38:45,000 --> 00:38:48,480 Speaker 1: actually care about technology. The Internet made me who I am, 741 00:38:48,520 --> 00:38:50,920 Speaker 1: connecting me and in many cases, introducing me to the 742 00:38:50,920 --> 00:38:53,239 Speaker 1: people I hold dearest to my heart. It let me 743 00:38:53,320 --> 00:38:57,440 Speaker 1: run a successful PR business despite having a learning disability dyspraxia, 744 00:38:57,520 --> 00:39:00,560 Speaker 1: or as it's called in America, developmental coordinational distort that 745 00:39:00,640 --> 00:39:02,360 Speaker 1: makes it difficult for me to write words with a 746 00:39:02,360 --> 00:39:05,560 Speaker 1: pen and thrive, despite being regularly told in secondary school 747 00:39:05,560 --> 00:39:07,879 Speaker 1: that I wouldn't amount to much any of you listen, then, guy, 748 00:39:07,920 --> 00:39:11,080 Speaker 1: fuck yourself. I believe that there are many, many, many 749 00:39:11,160 --> 00:39:13,480 Speaker 1: people that have been allowed to live better, full and 750 00:39:13,560 --> 00:39:16,840 Speaker 1: more meaningful lives as a result of technology, and I 751 00:39:16,880 --> 00:39:18,880 Speaker 1: believe that the people who built the tools that allowed 752 00:39:18,880 --> 00:39:21,560 Speaker 1: them to do so no longer control the future. And 753 00:39:21,640 --> 00:39:25,120 Speaker 1: that makes me really fucking annoyed. I refuse to watch 754 00:39:25,160 --> 00:39:28,000 Speaker 1: this bullshit happen in silence, and I encourage you, as 755 00:39:28,040 --> 00:39:31,319 Speaker 1: I have before, to fill social media and conversations with 756 00:39:31,360 --> 00:39:34,759 Speaker 1: the names of those responsible. Satch In a Dela of Microsoft, 757 00:39:34,920 --> 00:39:38,480 Speaker 1: Sundar Peshai of Google, Mark Zuckerberg of Meta, Andy Jesse 758 00:39:38,600 --> 00:39:43,239 Speaker 1: of Amazon, Sam Altman of Associated cons Mir Marati CTO 759 00:39:43,400 --> 00:39:46,719 Speaker 1: of Open Ai, and Read Hoffman, the venture capitalist who 760 00:39:46,760 --> 00:39:50,799 Speaker 1: seems to always benefit over the next bullshit machine, all 761 00:39:50,880 --> 00:39:55,360 Speaker 1: responsible for this echoing nihilism that's created the generative AI boom. 762 00:39:55,880 --> 00:39:57,759 Speaker 1: All of these people will be fine, by the way, 763 00:39:57,760 --> 00:40:01,000 Speaker 1: when this collapses. All of these people will just fine. 764 00:40:01,239 --> 00:40:04,560 Speaker 1: Multi billionaires, they will be just fine. But the people 765 00:40:04,560 --> 00:40:06,479 Speaker 1: that won't be are the people they will lay off, 766 00:40:06,640 --> 00:40:09,800 Speaker 1: the people that have invested and lost money in these things, 767 00:40:10,239 --> 00:40:12,719 Speaker 1: the people that may even come to rely on chat 768 00:40:12,760 --> 00:40:15,120 Speaker 1: GPT in some small way, who won't be able to 769 00:40:15,120 --> 00:40:17,320 Speaker 1: afford it when it goes tits up or gets jacked 770 00:40:17,400 --> 00:40:19,640 Speaker 1: up price wise, because that's the only way to make 771 00:40:19,680 --> 00:40:22,319 Speaker 1: it affordable for the company. Or maybe all of this 772 00:40:22,520 --> 00:40:26,640 Speaker 1: just disappears and these customers are left with nothing. But 773 00:40:26,760 --> 00:40:30,799 Speaker 1: sometimes I understand you might feel a little doomed, the 774 00:40:30,840 --> 00:40:33,439 Speaker 1: doomer in you. Someone recently called me a duma, which 775 00:40:33,760 --> 00:40:37,480 Speaker 1: really pissed me off, because I'm not I'm angry, and 776 00:40:37,560 --> 00:40:39,400 Speaker 1: I know some of you might be too. I won't 777 00:40:39,640 --> 00:40:41,960 Speaker 1: say you're angry, that isn't fair, but based on the 778 00:40:42,000 --> 00:40:45,120 Speaker 1: emails again, some of you seem pretty pissed. And I'm 779 00:40:45,120 --> 00:40:47,320 Speaker 1: not speaking for you here, but I'm angry because I 780 00:40:47,440 --> 00:40:50,120 Speaker 1: know that TED can be cooler and better and more interesting, 781 00:40:50,360 --> 00:40:53,480 Speaker 1: that there are real problems of poverty and climate change, 782 00:40:53,960 --> 00:40:57,960 Speaker 1: of the myriad systems that don't connect very well on 783 00:40:58,080 --> 00:41:02,320 Speaker 1: every laptop and phone, That things can be better, problems 784 00:41:02,360 --> 00:41:05,319 Speaker 1: that can be solved, but right now they need to 785 00:41:05,320 --> 00:41:09,279 Speaker 1: put tens of billions and dollars into generative AI. Come on, man, 786 00:41:09,320 --> 00:41:12,359 Speaker 1: we can do better than this. And while there might 787 00:41:12,400 --> 00:41:15,400 Speaker 1: be little you can do to them directly, I encourage 788 00:41:15,440 --> 00:41:18,080 Speaker 1: you to say these names to your friends on social media. 789 00:41:18,480 --> 00:41:21,560 Speaker 1: Prabagar Ragavan, remember him from Google. Tell people all about 790 00:41:21,640 --> 00:41:26,719 Speaker 1: him too. I encourage you to let anybody and everybody 791 00:41:26,760 --> 00:41:30,880 Speaker 1: you know know who's responsible for this, because ultimately what 792 00:41:31,040 --> 00:41:35,360 Speaker 1: may pop the AI bubble is a narrative, and narratives shift, 793 00:41:35,640 --> 00:41:38,480 Speaker 1: and narratives are shifted by word of mouth, and they 794 00:41:38,520 --> 00:41:41,000 Speaker 1: shift because of people like you and like me who 795 00:41:41,000 --> 00:41:51,719 Speaker 1: never shut up. Thank you for listening to better Offline. 796 00:41:51,880 --> 00:41:54,279 Speaker 2: The editor and composer of the Better Offline theme song 797 00:41:54,360 --> 00:41:57,000 Speaker 2: is Matasowski. You can check out more of his music 798 00:41:57,040 --> 00:42:00,680 Speaker 2: and audio projects at Matasowski dot com, M A T 799 00:42:00,680 --> 00:42:05,160 Speaker 2: T O S O W s ki dot com. You 800 00:42:05,160 --> 00:42:07,680 Speaker 2: can email me at easy at Better Offline dot com 801 00:42:07,760 --> 00:42:10,080 Speaker 2: or visit Better Offline dot com to find more podcast 802 00:42:10,120 --> 00:42:13,440 Speaker 2: links and of course, my newsletter. I also really recommend 803 00:42:13,480 --> 00:42:15,400 Speaker 2: you go to chat dot where's youreaed dot at to 804 00:42:15,520 --> 00:42:17,879 Speaker 2: visit the discord, and go to our slash. 805 00:42:17,560 --> 00:42:19,440 Speaker 1: Better Offline to check out our reddit. 806 00:42:20,200 --> 00:42:21,560 Speaker 2: Thank you so much for listening. 807 00:42:22,400 --> 00:42:25,080 Speaker 1: Better Offline is a production of cool Zone Media. For 808 00:42:25,200 --> 00:42:28,360 Speaker 1: more from cool Zone Media, visit our website cool Zonemedia 809 00:42:28,440 --> 00:42:31,279 Speaker 1: dot com, or check us out on the iHeartRadio app, 810 00:42:31,320 --> 00:42:34,000 Speaker 1: Apple Podcasts, or wherever you get your podcasts.