1 00:00:02,800 --> 00:00:03,560 Speaker 1: Zone Media. 2 00:00:04,880 --> 00:00:07,160 Speaker 2: Hello and welcome to Better Offline. I'm your host ed 3 00:00:07,240 --> 00:00:22,120 Speaker 2: zip tron. Check out the episode notes. Got a wonderful 4 00:00:22,160 --> 00:00:24,800 Speaker 2: merchandise and completely separate to the podcast, got a wonderful 5 00:00:24,800 --> 00:00:27,040 Speaker 2: newsletter Where's your Head dot at with the premium section 6 00:00:27,080 --> 00:00:30,360 Speaker 2: that I would love you to subscribe to. But I 7 00:00:30,400 --> 00:00:33,320 Speaker 2: also got some good news. You've got another three part episode, 8 00:00:33,320 --> 00:00:35,760 Speaker 2: the second of the year, and this week we're going 9 00:00:35,800 --> 00:00:37,839 Speaker 2: to be talking about how the cracks and the generative 10 00:00:37,840 --> 00:00:40,560 Speaker 2: AI market are becoming harder to ignore, and how recent 11 00:00:40,560 --> 00:00:43,000 Speaker 2: events are making a collapse seem all the more inevitable. 12 00:00:43,240 --> 00:00:45,840 Speaker 2: What that means for the wider economy, really the markets 13 00:00:46,400 --> 00:00:49,400 Speaker 2: and you and I want to make that case because 14 00:00:49,440 --> 00:00:51,440 Speaker 2: as a journalist, I believe I have the duty to 15 00:00:51,440 --> 00:00:53,400 Speaker 2: give you the information you need to make sense of 16 00:00:53,400 --> 00:00:58,560 Speaker 2: a world increasingly feeling incomprehensible and of course detachment reality, 17 00:00:58,600 --> 00:01:01,800 Speaker 2: and where everything is consequent to everyone, where it's impossible 18 00:01:01,840 --> 00:01:04,520 Speaker 2: for ordinary people to shroud themselves from the consequences of 19 00:01:04,600 --> 00:01:08,880 Speaker 2: decisions made by the executive and shareholder class. Good journalism 20 00:01:09,000 --> 00:01:11,679 Speaker 2: is making sure that history is actively captured and appropriately 21 00:01:11,760 --> 00:01:14,560 Speaker 2: described in the cest, and it's the accurate to describe 22 00:01:14,600 --> 00:01:18,040 Speaker 2: things as they currently are as alarming, and boy howdy 23 00:01:18,040 --> 00:01:18,960 Speaker 2: am I alarmed? 24 00:01:19,640 --> 00:01:19,800 Speaker 3: Now? 25 00:01:19,840 --> 00:01:23,080 Speaker 2: Alarm is not a state of weakness or belligerence in myopia, 26 00:01:23,160 --> 00:01:25,520 Speaker 2: My concern does not dull my vision, even though it's 27 00:01:25,560 --> 00:01:29,120 Speaker 2: a convenient to frame me as somehow alarmist like I 28 00:01:29,160 --> 00:01:32,280 Speaker 2: have some hidden agenda or bias toward doom. I profoundly 29 00:01:32,280 --> 00:01:36,319 Speaker 2: dislike the financial waste, the environmental destruction, and fundamentally I 30 00:01:36,400 --> 00:01:39,200 Speaker 2: dislike the attempt to gaslight people into swearing fealty to 31 00:01:39,240 --> 00:01:42,240 Speaker 2: a sickly in frail pseudo industry where everybody but in 32 00:01:42,319 --> 00:01:46,120 Speaker 2: Nvidia and consultancies lose money. And I also dislike the 33 00:01:46,200 --> 00:01:48,080 Speaker 2: fact that I and others like me are held to 34 00:01:48,120 --> 00:01:51,440 Speaker 2: a remarkably different standard to those that paint themselves as optimists, 35 00:01:51,480 --> 00:01:54,800 Speaker 2: which typically means people that agree with what the market 36 00:01:54,840 --> 00:01:59,000 Speaker 2: wishes were true. Critics are continually badgered, prodded, poked, mocked, 37 00:01:59,000 --> 00:02:00,880 Speaker 2: and jeered out for not all the metrically aligning with 38 00:02:00,920 --> 00:02:03,560 Speaker 2: the idea that generative AI will be this massive industry, 39 00:02:03,800 --> 00:02:06,880 Speaker 2: constantly having to prove themselves as if somehow there's something 40 00:02:06,960 --> 00:02:10,480 Speaker 2: malevolent or craving about criticism. The critics do this for clicks, 41 00:02:10,560 --> 00:02:13,640 Speaker 2: or to be contrarian. I don't do anything for clicks 42 00:02:13,680 --> 00:02:16,360 Speaker 2: or downloads or prestige. I don't have any stocks or 43 00:02:16,400 --> 00:02:19,800 Speaker 2: short positions. My gender is simple. I like talking about 44 00:02:19,800 --> 00:02:21,720 Speaker 2: this crap and it comes to me naturally. I have 45 00:02:21,760 --> 00:02:24,079 Speaker 2: a podcast and it is on some level my job 46 00:02:24,120 --> 00:02:26,320 Speaker 2: to try and understand what the tech industry is doing 47 00:02:26,400 --> 00:02:26,919 Speaker 2: day to day. 48 00:02:27,919 --> 00:02:28,480 Speaker 3: And I get it. 49 00:02:28,800 --> 00:02:30,680 Speaker 2: I get It's easy to try and dismiss what I 50 00:02:30,720 --> 00:02:33,200 Speaker 2: say is going against the green because AI is big, 51 00:02:33,400 --> 00:02:35,520 Speaker 2: and AI is something we should all be impressed by, 52 00:02:35,720 --> 00:02:37,840 Speaker 2: and the AI is the thing that's going to start everything, 53 00:02:37,840 --> 00:02:40,120 Speaker 2: and all this fucking money is tied up in it. 54 00:02:41,120 --> 00:02:44,240 Speaker 2: But look, this isn't a fad for me. This isn't 55 00:02:44,280 --> 00:02:47,079 Speaker 2: something I'm doing because I feel like it because I'm 56 00:02:47,160 --> 00:02:50,680 Speaker 2: jumping to the next trend. No. I've been roiling against 57 00:02:50,720 --> 00:02:53,440 Speaker 2: bullshit bubble since twenty twenty one, the anti remote work 58 00:02:53,480 --> 00:02:55,600 Speaker 2: push and the people behind it, the clubhouse and audio 59 00:02:55,639 --> 00:02:58,480 Speaker 2: social networks bubble, the NFT bubble that made up quiet, 60 00:02:58,560 --> 00:03:01,919 Speaker 2: quitting panic, and even even and this one I got 61 00:03:01,919 --> 00:03:02,600 Speaker 2: no credit for. 62 00:03:03,040 --> 00:03:04,400 Speaker 3: I called that. 63 00:03:04,360 --> 00:03:06,800 Speaker 2: Something was up with the FTX several months before it imploded. 64 00:03:07,080 --> 00:03:08,960 Speaker 2: Did I do much more than that. No, I found 65 00:03:08,960 --> 00:03:13,760 Speaker 2: one thing. Nevertheless, this is in contrarianism, not at all. 66 00:03:13,840 --> 00:03:16,239 Speaker 2: It's the kind of skepticism of power and capital that's 67 00:03:16,280 --> 00:03:18,960 Speaker 2: necessary to meet these moments. And if it's necessary to 68 00:03:19,000 --> 00:03:21,519 Speaker 2: dismiss my work because it makes you feel icky inside, 69 00:03:21,760 --> 00:03:25,520 Speaker 2: get a therapist or see a priest. Nevertheless, I'm alarmed. 70 00:03:25,520 --> 00:03:27,520 Speaker 2: And while I have said some of these things separately, 71 00:03:27,560 --> 00:03:30,600 Speaker 2: based on recent developments, I think it's necessary to say why. 72 00:03:31,160 --> 00:03:33,720 Speaker 2: In short, I believe the AI bubble is deeply unstable, 73 00:03:33,720 --> 00:03:35,720 Speaker 2: built on vibes and blind faith. And when I say 74 00:03:35,760 --> 00:03:38,920 Speaker 2: the AI bubble, I mean the entirety of the AI trade. 75 00:03:39,160 --> 00:03:43,200 Speaker 2: And it's alarmingly simple too, he says, before doing three 76 00:03:43,400 --> 00:03:46,080 Speaker 2: episodes on it. But this isn't going to be some saccerin, 77 00:03:46,160 --> 00:03:49,280 Speaker 2: whiny or simply worrisome podcast. I think at this point 78 00:03:49,320 --> 00:03:51,360 Speaker 2: it's become a little ridiculous to not see we're in 79 00:03:51,440 --> 00:03:53,880 Speaker 2: a bubble. We are in a goddamn bubble. By the way, 80 00:03:53,920 --> 00:03:55,560 Speaker 2: it's so obvious we're in a bubble. It's been so 81 00:03:55,640 --> 00:03:57,880 Speaker 2: obvious we're in a bubble. It's been obvious for months, 82 00:03:57,880 --> 00:03:59,680 Speaker 2: if not years, A bubble that seems so strong, but 83 00:03:59,680 --> 00:04:02,840 Speaker 2: it's very weak with a central point of failure. I 84 00:04:02,880 --> 00:04:05,040 Speaker 2: may not be a contrarian, but I am a hater. 85 00:04:05,560 --> 00:04:07,760 Speaker 2: I hate the waste, the loss, the destruction, the theft, 86 00:04:07,760 --> 00:04:09,600 Speaker 2: the damage to our planet, and the sheer excitement. And 87 00:04:09,720 --> 00:04:13,040 Speaker 2: some executives and yes, some writers have that workers may 88 00:04:13,080 --> 00:04:16,200 Speaker 2: be replaced by AI and the bold faced fucking lie 89 00:04:16,240 --> 00:04:19,360 Speaker 2: that it's actually happening. And what Generative AI is doing 90 00:04:19,600 --> 00:04:22,599 Speaker 2: is somehow proof that it will. And so I present 91 00:04:22,680 --> 00:04:25,919 Speaker 2: to you the Hater's Guide to the AI Bubble, a 92 00:04:25,960 --> 00:04:28,559 Speaker 2: comprehensive rundown of the arguments I have against the current 93 00:04:28,560 --> 00:04:31,760 Speaker 2: AI booms existence. Send this podcast to your friends you 94 00:04:31,839 --> 00:04:34,240 Speaker 2: love ones or I don't know, blare it in their 95 00:04:34,320 --> 00:04:37,440 Speaker 2: ears like you're torturing them. But no, this isn't going 96 00:04:37,480 --> 00:04:39,839 Speaker 2: to be a traditional guy but something you can listen 97 00:04:39,920 --> 00:04:42,000 Speaker 2: to and say, oh, that's why the AI bubble is 98 00:04:42,000 --> 00:04:44,000 Speaker 2: so bad. And at this point, I know, I'm tired 99 00:04:44,040 --> 00:04:46,040 Speaker 2: of being gas lit by guys in Gingham shirts who 100 00:04:46,040 --> 00:04:48,360 Speaker 2: desperately want to curry favor with other guys in Gingham 101 00:04:48,400 --> 00:04:51,680 Speaker 2: shirts but who also have PhDs. I'm tired of hearing 102 00:04:51,720 --> 00:04:54,080 Speaker 2: people talk about how we're in the era of agents 103 00:04:54,080 --> 00:04:56,680 Speaker 2: that don't fucking work and will never fucking work. I'm 104 00:04:56,680 --> 00:04:59,120 Speaker 2: tired of hearing about powerful AI that's actually crap, and 105 00:04:59,120 --> 00:05:01,120 Speaker 2: I'm tired of being told the future is here while 106 00:05:01,120 --> 00:05:04,159 Speaker 2: having the world's least useful, most expensive cloud software shoved 107 00:05:04,160 --> 00:05:06,840 Speaker 2: down my throat and up my asshole. Look, the general 108 00:05:06,960 --> 00:05:09,360 Speaker 2: FAI boom is a mirage. It hasn't got the revenue 109 00:05:09,440 --> 00:05:11,599 Speaker 2: or the returns of the product efficacy for it to matter. 110 00:05:11,960 --> 00:05:15,200 Speaker 2: Everything you're seeing is ridiculous and wasteful when it all 111 00:05:15,240 --> 00:05:16,960 Speaker 2: goes tits up. And want you to remember that I 112 00:05:17,000 --> 00:05:19,760 Speaker 2: said this. I tried to say something. I've been trying 113 00:05:19,760 --> 00:05:22,880 Speaker 2: to say something for a while. But let's start with 114 00:05:23,560 --> 00:05:26,440 Speaker 2: something real obvious. Let's start by talking about the so 115 00:05:26,600 --> 00:05:29,320 Speaker 2: called Magnificent Seven's week point and it's not the one 116 00:05:29,320 --> 00:05:32,080 Speaker 2: you think because it's in video. As I write the 117 00:05:32,080 --> 00:05:34,520 Speaker 2: script for this podcast, in Vidia's sitting around one hundred 118 00:05:34,520 --> 00:05:36,880 Speaker 2: and seventy bucks a share, a dramatic reversal of faith 119 00:05:36,920 --> 00:05:39,120 Speaker 2: after the pummeling it took from the deep Seek situation 120 00:05:39,240 --> 00:05:41,720 Speaker 2: in January, which sent it tumbling to a brief late 121 00:05:41,720 --> 00:05:44,800 Speaker 2: April trip below one hundred dollars. Four things turned around 122 00:05:45,960 --> 00:05:51,080 Speaker 2: the mag seven stocks and Video, Microsoft, Alphabet, which is Google, Apple, Meta, Tesla, 123 00:05:51,120 --> 00:05:53,520 Speaker 2: and Amazon make up around thirty five percent of the 124 00:05:53,560 --> 00:05:55,600 Speaker 2: value of the US stock market, and of that and 125 00:05:55,720 --> 00:05:58,560 Speaker 2: Video's market value takes up about nineteen percent of the 126 00:05:58,560 --> 00:06:01,840 Speaker 2: Magnificent seven, eight to nine percent of the entire US 127 00:06:01,880 --> 00:06:05,160 Speaker 2: stock market. It's not brilliant. This dominance is also why 128 00:06:05,320 --> 00:06:08,040 Speaker 2: ordinary people ought to be deeply concerned about the AI bubble. 129 00:06:08,160 --> 00:06:10,400 Speaker 2: The Magnificent seven is almost certainly a big part of 130 00:06:10,440 --> 00:06:13,880 Speaker 2: their retirement plans, even if they're not directly invested. Back 131 00:06:13,920 --> 00:06:17,400 Speaker 2: in May, the Wonderful Laura Branton from Yahoo Finance reported 132 00:06:17,440 --> 00:06:21,040 Speaker 2: the Microsoft, Amazon, Meta, Alphabet, and Tesla alone make up 133 00:06:21,080 --> 00:06:25,400 Speaker 2: forty two point four percent of Invidia's revenue. The breakdown 134 00:06:25,560 --> 00:06:29,039 Speaker 2: doesn't make things better. Meta spends twenty five percent, and 135 00:06:29,160 --> 00:06:32,240 Speaker 2: Microsoft on alarming forty seven percent of their capital expenditures 136 00:06:32,279 --> 00:06:35,280 Speaker 2: and on Nvidia chips and as Branton notes, Microsoft also 137 00:06:35,320 --> 00:06:38,880 Speaker 2: spends money renting service from core Weave, which analyst Guild 138 00:06:39,000 --> 00:06:42,320 Speaker 2: Luria of DA Davidson estimates accounted for eight billion dollars 139 00:06:42,400 --> 00:06:45,120 Speaker 2: more than six percent of Invidia's revenue in twenty twenty four. 140 00:06:46,480 --> 00:06:49,480 Speaker 2: Laurier also estimates that neo cloud companies like Coreweav and 141 00:06:49,520 --> 00:06:52,880 Speaker 2: Crusoe that exist only to provide AI compute services account 142 00:06:52,880 --> 00:06:55,000 Speaker 2: for as much as ten percent of Nvidia's revenue, or 143 00:06:55,040 --> 00:06:57,760 Speaker 2: at least did so in twenty twenty four. In vidious 144 00:06:57,800 --> 00:07:00,840 Speaker 2: climbing stock value comes from one thing. It's continued revenue 145 00:07:00,839 --> 00:07:03,640 Speaker 2: growth in the past four quarters. And video has seen 146 00:07:03,720 --> 00:07:05,800 Speaker 2: year of a year growth of one hundred and one percent, 147 00:07:05,880 --> 00:07:08,520 Speaker 2: ninety four percent, seventy eight percent, and sixty nine percent. 148 00:07:08,720 --> 00:07:11,200 Speaker 2: And in the last quarter a little statistic was carefully 149 00:07:11,240 --> 00:07:15,760 Speaker 2: brushed under the rug in Video missed though narrowly, on 150 00:07:15,880 --> 00:07:19,000 Speaker 2: data center revenue. And data center revenue is, by the way, 151 00:07:19,040 --> 00:07:22,640 Speaker 2: where the GPUs go, and all the associated hardware and 152 00:07:22,080 --> 00:07:25,680 Speaker 2: so kind of server architecture switches and the like, and yeah, 153 00:07:25,720 --> 00:07:27,720 Speaker 2: this is exactly what it sounds like. GPUs that are 154 00:07:27,800 --> 00:07:30,800 Speaker 2: used in servers rather than gaming consoles and PCs. I 155 00:07:30,840 --> 00:07:32,480 Speaker 2: get a lot of emails saying, oh, will it be 156 00:07:32,520 --> 00:07:34,960 Speaker 2: easier for me to buy consumer graphics cards? I don't 157 00:07:35,000 --> 00:07:38,680 Speaker 2: fucking know, mate, I'm just did to talk about enterprise bullshit. Okay, 158 00:07:38,720 --> 00:07:41,760 Speaker 2: not really enterprise, but enterprise scale GPUs were getting off track. 159 00:07:42,360 --> 00:07:45,480 Speaker 2: Analyst estimated it would make. It would make thirty nine 160 00:07:45,520 --> 00:07:47,920 Speaker 2: point four billion dollars from the data center category, and 161 00:07:47,960 --> 00:07:51,480 Speaker 2: then video only only. I no pathetic amount brought in 162 00:07:51,600 --> 00:07:55,200 Speaker 2: thirty nine point one billion dollars. Then again, this could 163 00:07:55,240 --> 00:07:57,480 Speaker 2: be attributed to their problems in China, especially as the 164 00:07:57,600 --> 00:08:00,600 Speaker 2: H twenty ban they were banned from selling uspace chip 165 00:08:00,640 --> 00:08:03,640 Speaker 2: in China has only just been lifted. In any case, 166 00:08:03,680 --> 00:08:06,200 Speaker 2: this was a miss, and I'm not sure why no 167 00:08:06,240 --> 00:08:07,680 Speaker 2: one wanted to talk about them. 168 00:08:08,520 --> 00:08:10,280 Speaker 3: But there's another problems. 169 00:08:10,440 --> 00:08:13,040 Speaker 2: There's so many little problems here, like in videos. Quarter 170 00:08:13,120 --> 00:08:16,680 Speaker 2: over quarter growth has also become aggressively normal. It went 171 00:08:16,680 --> 00:08:19,120 Speaker 2: from sixty nine percent to fifty nine percent to twelve 172 00:08:19,160 --> 00:08:23,120 Speaker 2: percent to twelve percent again a quarter of quita, which 173 00:08:23,240 --> 00:08:26,320 Speaker 2: again isn't bad, it's pretty great in fact, But when 174 00:08:26,560 --> 00:08:28,680 Speaker 2: eighty eight percent of your revenue is based on one 175 00:08:28,680 --> 00:08:31,840 Speaker 2: particular line in your earnings, it's a pretty big concern, 176 00:08:31,920 --> 00:08:34,840 Speaker 2: at least for me. Look, I'm no stock analysts do 177 00:08:34,920 --> 00:08:37,559 Speaker 2: not take stock advice from me. I don't know about stocks. 178 00:08:37,559 --> 00:08:39,080 Speaker 2: I don't know what will go up and down, but 179 00:08:39,120 --> 00:08:42,079 Speaker 2: I'll tell you, I'll tell you something's not right here. 180 00:08:42,320 --> 00:08:44,720 Speaker 2: But I'm going to keep this simple in Video relies 181 00:08:44,720 --> 00:08:47,400 Speaker 2: on not only selling lots of GPUs each quarter, but 182 00:08:47,440 --> 00:08:50,560 Speaker 2: it must always sell more of them the following quarter. 183 00:08:51,600 --> 00:08:54,200 Speaker 2: More than forty two percent of revenue, and Video's revenue 184 00:08:54,200 --> 00:08:57,840 Speaker 2: comes from Microsoft, Amazon, Meta Alphabet, and Tesla continuing to 185 00:08:58,000 --> 00:09:02,640 Speaker 2: buy more GPUs. Remember, it's not about buying the same amount. 186 00:09:02,920 --> 00:09:07,120 Speaker 2: Number must go up. In Vidia's continued value and continued 187 00:09:07,120 --> 00:09:10,560 Speaker 2: growth is heavily reliant on hyperscalar purchases and continued interest 188 00:09:10,600 --> 00:09:13,720 Speaker 2: in generative AI. But really just the buying part the 189 00:09:13,760 --> 00:09:17,319 Speaker 2: GPUs are what matter. And the US stock markets continued 190 00:09:17,360 --> 00:09:19,960 Speaker 2: health relies on some level of five or six companies, 191 00:09:19,960 --> 00:09:23,600 Speaker 2: and it's unclear how many GPUs Apple buys, spending billions 192 00:09:23,640 --> 00:09:26,439 Speaker 2: of dollars on GPUs from Nvidia and more every quarter. 193 00:09:26,640 --> 00:09:30,880 Speaker 2: In fact, I found an analysis from portfolio manager Don 194 00:09:30,920 --> 00:09:33,480 Speaker 2: ke Wang from January twenty twenty five. The found that 195 00:09:33,480 --> 00:09:36,360 Speaker 2: the Magnificent seven stocks accounted for forty seven point eighty 196 00:09:36,440 --> 00:09:39,400 Speaker 2: seven percent of the Russell one thousand indexes returns in 197 00:09:39,440 --> 00:09:41,520 Speaker 2: twenty twenty four. And that's an index fund of the 198 00:09:41,880 --> 00:09:45,120 Speaker 2: thousand highest ranked stocks on the foot Sea Russell's Index, 199 00:09:45,240 --> 00:09:48,160 Speaker 2: which in simpler terms means thirty five percent of the 200 00:09:48,240 --> 00:09:50,040 Speaker 2: US stock market is held up by five or six 201 00:09:50,080 --> 00:09:54,240 Speaker 2: companies buying GPUs. If a video's growth story stumbles, it 202 00:09:54,280 --> 00:09:56,400 Speaker 2: will reverberate through the rest of the mag seven to two, 203 00:09:56,880 --> 00:10:00,960 Speaker 2: making them rely on their own AI trade stories. And 204 00:10:01,000 --> 00:10:03,240 Speaker 2: you know it when you know it, when you look 205 00:10:03,240 --> 00:10:05,760 Speaker 2: at the stories. There is no AI trade because Generative 206 00:10:05,760 --> 00:10:08,440 Speaker 2: AI is not making anybody any goddamn money. But I 207 00:10:08,480 --> 00:10:10,600 Speaker 2: have to make money, which is why you need to 208 00:10:10,640 --> 00:10:13,079 Speaker 2: listen to the following advertisement. It's the only one of 209 00:10:13,080 --> 00:10:26,320 Speaker 2: these bloody things you're gonna get. Here's an ad and 210 00:10:26,440 --> 00:10:30,000 Speaker 2: we're back. And I am so tired of people telling 211 00:10:30,080 --> 00:10:32,319 Speaker 2: me that companies are making tons of money on AI. 212 00:10:32,600 --> 00:10:35,360 Speaker 2: They are not. Anyone saying this to you is lying 213 00:10:35,520 --> 00:10:39,040 Speaker 2: or ignorant or both. The Magnificent Seven spent an insane 214 00:10:39,080 --> 00:10:41,920 Speaker 2: five hundred and sixty billion dollars between twenty twenty four 215 00:10:41,920 --> 00:10:45,160 Speaker 2: and twenty twenty five on Capex, with the overwhelming majority 216 00:10:45,200 --> 00:10:49,079 Speaker 2: going towards Generative AI and their effort for this wonderful effort. 217 00:10:49,160 --> 00:10:52,000 Speaker 2: For that more than half a trillion dollars, these companies 218 00:10:52,000 --> 00:10:54,520 Speaker 2: have made about thirty five billion dollars in revenue and 219 00:10:54,600 --> 00:10:57,640 Speaker 2: no profit. And I must say, and this is a 220 00:10:57,679 --> 00:11:01,520 Speaker 2: technical term, this is egregiously fucking stud But let's break 221 00:11:01,559 --> 00:11:04,200 Speaker 2: it down. Start with starting out in Redmond with our 222 00:11:04,240 --> 00:11:07,319 Speaker 2: friends at Microsoft, and they plan to spend eighty billion 223 00:11:07,360 --> 00:11:10,800 Speaker 2: dollars on CAPEX in twenty twenty five. Now, mustn't. January 224 00:11:10,840 --> 00:11:14,720 Speaker 2: twenty twenty five, Microsoft's annualized revenue meaning best month times 225 00:11:14,760 --> 00:11:18,000 Speaker 2: twelve from our official intelligence, was thirteen billion dollars, a 226 00:11:18,080 --> 00:11:21,720 Speaker 2: number that it's chosen, not what they since slightly because 227 00:11:21,760 --> 00:11:24,400 Speaker 2: said number is either flat or not growing, though it 228 00:11:24,480 --> 00:11:26,280 Speaker 2: could in its upcoming I think at the end of 229 00:11:26,320 --> 00:11:29,360 Speaker 2: this week or next one, it's got cut earnings coming up. 230 00:11:29,400 --> 00:11:31,200 Speaker 3: Maybe they'd be good news. Yeah. 231 00:11:31,200 --> 00:11:33,080 Speaker 2: The problem with this revenue is that ten billion dollars 232 00:11:33,120 --> 00:11:35,440 Speaker 2: of that revenue, according to the information, comes from open 233 00:11:35,480 --> 00:11:38,760 Speaker 2: AIS spend on a Microsoft' as or cloud, and Microsoft 234 00:11:38,800 --> 00:11:42,280 Speaker 2: offers preferential pricing. I'm quoting the information here at a 235 00:11:42,280 --> 00:11:45,560 Speaker 2: heavily discounted rental rate that essentially only covers Microsoft's costs 236 00:11:45,559 --> 00:11:51,320 Speaker 2: for operating the servers. That's not good, right, like it's 237 00:11:51,360 --> 00:11:54,319 Speaker 2: not good. It's not good that seventy six point nine 238 00:11:54,360 --> 00:11:57,520 Speaker 2: percent of Microsoft's AI revenue comes from open AI, and 239 00:11:57,920 --> 00:12:00,920 Speaker 2: that revenue is made at cost or just above it, 240 00:12:01,160 --> 00:12:04,280 Speaker 2: which makes Microsoft's real AI revenue about three billion dollars 241 00:12:04,400 --> 00:12:06,800 Speaker 2: or about three point seven five percent of this year's 242 00:12:06,800 --> 00:12:10,199 Speaker 2: capital expenditures, or sixteen point two five percent if you 243 00:12:10,240 --> 00:12:13,200 Speaker 2: count open AIS revenue, which costs Microsoft likely more money 244 00:12:13,240 --> 00:12:16,640 Speaker 2: than it earns. The information also reports that Microsoft made 245 00:12:16,720 --> 00:12:19,800 Speaker 2: four point seven billion dollars in AI revenue in twenty 246 00:12:19,840 --> 00:12:22,640 Speaker 2: twenty four, of which open AI accounted for two billion dollars, 247 00:12:22,880 --> 00:12:25,199 Speaker 2: meaning that for the one hundred and thirty five point 248 00:12:25,240 --> 00:12:27,720 Speaker 2: seven billion dollars that Microsoft are spent in two years 249 00:12:27,760 --> 00:12:32,840 Speaker 2: in AI infrastructure, it's made seventeen point seven billion dollars, 250 00:12:32,840 --> 00:12:36,679 Speaker 2: of which open AI was twelve point seven billion dollars. 251 00:12:37,200 --> 00:12:39,880 Speaker 2: It's kind of crap, isn't it. It's not very good 252 00:12:39,920 --> 00:12:42,080 Speaker 2: at all, and things do not improve when we get 253 00:12:42,120 --> 00:12:45,640 Speaker 2: to Amazon. An analyst estimates that Amazon, which plans to 254 00:12:45,679 --> 00:12:48,120 Speaker 2: spend one hundred and five billion dollars in capital expenditures 255 00:12:48,120 --> 00:12:50,680 Speaker 2: this year, will make five billion dollars in AI in 256 00:12:50,720 --> 00:12:53,040 Speaker 2: twenty twenty five, rising and I quote as much as 257 00:12:53,040 --> 00:12:55,880 Speaker 2: eighty percent, suggesting that Amazon might have made a measly 258 00:12:55,880 --> 00:12:58,319 Speaker 2: two point seven seven billion dollars in twenty twenty four 259 00:12:58,360 --> 00:13:00,559 Speaker 2: on AI in a year when it's spent eighty three 260 00:13:00,640 --> 00:13:04,400 Speaker 2: billion dollars in capital expenditures, and last year, Amazon CEO 261 00:13:04,480 --> 00:13:07,360 Speaker 2: Andy Jesse said that and I quote, AI represents for 262 00:13:07,400 --> 00:13:09,800 Speaker 2: sure the biggest opportunity since cloud and probably the biggest 263 00:13:09,840 --> 00:13:13,320 Speaker 2: technologies shift an opportunity in business since the Internet. I 264 00:13:13,360 --> 00:13:16,280 Speaker 2: personally think he is full of shit. And it's a 265 00:13:16,320 --> 00:13:19,000 Speaker 2: similar story over with Google, which plans to spend seventy 266 00:13:19,040 --> 00:13:22,920 Speaker 2: five billion dollars in capex twenty twenty five. Bank of 267 00:13:22,960 --> 00:13:25,720 Speaker 2: America analysts justin Post estimated a few weeks ago that 268 00:13:25,760 --> 00:13:27,840 Speaker 2: Google's AI revenue would be in the region of seven 269 00:13:27,840 --> 00:13:30,880 Speaker 2: point seven billion dollars, though his math, if I'm honest, 270 00:13:31,000 --> 00:13:33,520 Speaker 2: is a little generous because it includes subscribers to packages 271 00:13:33,559 --> 00:13:35,080 Speaker 2: that include a lot of non AI staff. 272 00:13:35,120 --> 00:13:35,280 Speaker 3: Two. 273 00:13:36,000 --> 00:13:40,320 Speaker 2: Google's one subscription includes increased cloud storage across Google Drive, Gmail, 274 00:13:40,320 --> 00:13:42,800 Speaker 2: and Google Photos, and added a twenty dollars a month 275 00:13:42,840 --> 00:13:46,760 Speaker 2: premium plan in February twenty twenty four that included access 276 00:13:46,800 --> 00:13:50,440 Speaker 2: to Google's various AI models. Google's claim that the premium 277 00:13:50,480 --> 00:13:52,520 Speaker 2: AI tier accounts some millions of the one hundred and 278 00:13:52,559 --> 00:13:55,679 Speaker 2: fifty million subscribers to Google one, though how many millions 279 00:13:55,720 --> 00:13:58,640 Speaker 2: is impossible to estimate That one would stop me trying, though, 280 00:13:58,840 --> 00:14:01,200 Speaker 2: Assuming the three point one billion dollars in twenty twenty 281 00:14:01,240 --> 00:14:03,520 Speaker 2: five revenue would work out to two hundred and fifty 282 00:14:03,520 --> 00:14:05,240 Speaker 2: eight million dollars a month, that would mean there were 283 00:14:05,320 --> 00:14:08,560 Speaker 2: twelve point nine million Google one subscribers also paying for 284 00:14:08,600 --> 00:14:11,800 Speaker 2: the premium Ai tier. This isn't out of the realm 285 00:14:11,880 --> 00:14:15,000 Speaker 2: of possibility, after all, Open ai has like fifteen point 286 00:14:15,080 --> 00:14:17,960 Speaker 2: five million paying subscribers, but post is making in a 287 00:14:18,320 --> 00:14:21,520 Speaker 2: kind of a generous assumption here. Nevertheless, well accept the 288 00:14:21,600 --> 00:14:25,080 Speaker 2: numbers as they are, because they fucking stink. Google's one 289 00:14:25,080 --> 00:14:28,760 Speaker 2: point one billion dollar in workspace revenue came from a 290 00:14:28,800 --> 00:14:31,640 Speaker 2: forced price hike on those who use Google services to 291 00:14:31,680 --> 00:14:34,760 Speaker 2: run their businesses. My ass included meaning that it's not 292 00:14:34,920 --> 00:14:37,200 Speaker 2: likely a number that they can significantly increase in the 293 00:14:37,200 --> 00:14:39,680 Speaker 2: future because it was just raising the rent on everyone. 294 00:14:40,000 --> 00:14:42,840 Speaker 2: And that's seven point seven billion dollars of revenue, not 295 00:14:42,960 --> 00:14:47,360 Speaker 2: profit on seventy five billion dollars of capital expenditures. Very nasty. 296 00:14:48,080 --> 00:14:50,680 Speaker 2: But let's move on to one of my faves, Meta, 297 00:14:50,760 --> 00:14:53,560 Speaker 2: which plans to spend seventy two billion dollars in twenty 298 00:14:53,600 --> 00:14:56,160 Speaker 2: twenty five. Someone's going to get mad at me for 299 00:14:56,240 --> 00:14:58,640 Speaker 2: saying this, But I believe that Meta is simply burning 300 00:14:58,720 --> 00:15:02,240 Speaker 2: cash on generative. There is no product the Meta sells 301 00:15:02,280 --> 00:15:05,560 Speaker 2: that monetizes large language buddles that I can tell at least, 302 00:15:05,560 --> 00:15:08,520 Speaker 2: But every Meta product now has them kind of shoved 303 00:15:08,520 --> 00:15:11,680 Speaker 2: in there. Year Instagram dms oinking at you to generate 304 00:15:11,800 --> 00:15:17,120 Speaker 2: artwork based in your conversation. Nevertheless, they do make some money, allegedly, 305 00:15:17,160 --> 00:15:19,120 Speaker 2: and we do have some sort of knowledge of what 306 00:15:19,280 --> 00:15:22,520 Speaker 2: Meta is saying they make due to a copyright infringement 307 00:15:22,600 --> 00:15:26,520 Speaker 2: case cardre versus Meta on sealed judgment briefs revealed in 308 00:15:26,560 --> 00:15:29,280 Speaker 2: April that Meta is claiming that Jenai driven revenue will 309 00:15:29,320 --> 00:15:31,160 Speaker 2: be more than two billion dollars in this year, with 310 00:15:31,320 --> 00:15:34,600 Speaker 2: estimates as high as three billion dollars. The same document 311 00:15:34,640 --> 00:15:36,840 Speaker 2: also claims that Meta expects to make four hundred and 312 00:15:36,920 --> 00:15:39,520 Speaker 2: sixty billion to one point four trillion dollars in total 313 00:15:39,560 --> 00:15:43,040 Speaker 2: revenue through twenty thirty five. And this is from Ai, 314 00:15:43,120 --> 00:15:45,040 Speaker 2: by the way, And this is the kind of thing 315 00:15:45,120 --> 00:15:47,120 Speaker 2: that should have you wrenched out of them. But you 316 00:15:47,120 --> 00:15:49,520 Speaker 2: should your key God should stop working when the words 317 00:15:49,600 --> 00:15:52,480 Speaker 2: leave your mouth because Meta makes ninety nine percent of 318 00:15:52,480 --> 00:15:55,200 Speaker 2: its revenue from advertising, and the unsealed documents state that 319 00:15:55,280 --> 00:15:58,240 Speaker 2: it generates from its Larmer models and will continue earning 320 00:15:58,280 --> 00:16:01,120 Speaker 2: revenue from each iteration and share percentage of the revenue 321 00:16:01,120 --> 00:16:03,520 Speaker 2: it generates from users of the Lama models hosted by 322 00:16:03,600 --> 00:16:07,160 Speaker 2: those companies. But the companies in question redacted. Mister Max 323 00:16:07,240 --> 00:16:10,800 Speaker 2: Zeph of tech runchads that metaalist host partners like Amazon 324 00:16:10,840 --> 00:16:15,000 Speaker 2: Web Services and Video Data Breaks, Grock Dell, Microsoft, as Yure, 325 00:16:15,080 --> 00:16:17,880 Speaker 2: Google Cloud, and Snowflake, so it's possible that Meta makes 326 00:16:18,040 --> 00:16:22,520 Speaker 2: money licensing to those companies. Sadly, the exhibits further discussing 327 00:16:22,560 --> 00:16:25,440 Speaker 2: these numbers are filed on the seal and also their 328 00:16:25,520 --> 00:16:29,680 Speaker 2: large language model is open source. What service is Meta providing? 329 00:16:29,960 --> 00:16:32,560 Speaker 2: Are these companies so goddamn lazy that they need Meta 330 00:16:32,600 --> 00:16:36,640 Speaker 2: to come in and set up the fruit? And Jesus Christ, 331 00:16:36,880 --> 00:16:39,120 Speaker 2: Jesus Christ, when I read these numbers, I just when 332 00:16:39,160 --> 00:16:42,400 Speaker 2: I read about these people who drive me a little insane. 333 00:16:42,800 --> 00:16:44,880 Speaker 2: Either way, we are now at three hundred and thirty 334 00:16:44,880 --> 00:16:47,560 Speaker 2: two billion dollars of capital expenditures in twenty twenty five. 335 00:16:47,600 --> 00:16:49,800 Speaker 2: For twenty eight point seven billion dollars of revenue, off 336 00:16:49,800 --> 00:16:51,880 Speaker 2: which ten billion dollars of it is open AI is 337 00:16:51,920 --> 00:16:55,360 Speaker 2: at cost or just above cost. Revenue not great. Then 338 00:16:55,400 --> 00:16:58,320 Speaker 2: there's Tesla, which doesn't appear to make money from generative 339 00:16:58,360 --> 00:17:00,760 Speaker 2: AI and plans to spend eleven billion dollars on capex 340 00:17:00,760 --> 00:17:04,399 Speaker 2: in twenty twenty five. Despite its media prominence in the 341 00:17:04,440 --> 00:17:06,720 Speaker 2: Magnificent seven. At least, Tesla is one of the least 342 00:17:06,760 --> 00:17:09,600 Speaker 2: exposed companies of mag seven to the AI trade, as 343 00:17:09,600 --> 00:17:12,080 Speaker 2: Elon Musk has turned it into a memestock company where 344 00:17:12,119 --> 00:17:14,920 Speaker 2: what they do doesn't really matter. That doesn't mean, of course, 345 00:17:14,960 --> 00:17:17,760 Speaker 2: that Musk isn't touching AI. The XAI, the company that 346 00:17:17,840 --> 00:17:20,760 Speaker 2: develops racist large language model grok and owns what remains 347 00:17:20,760 --> 00:17:23,439 Speaker 2: of Twitter, apparently burns a billion dollars a month, and 348 00:17:23,480 --> 00:17:25,960 Speaker 2: The Information reports that it makes a whopping hundred million 349 00:17:26,000 --> 00:17:28,800 Speaker 2: dollars of annualized revenue, so about eight point three three 350 00:17:28,880 --> 00:17:32,159 Speaker 2: million dollars a month. Now there's a shareholder vote for 351 00:17:32,200 --> 00:17:35,280 Speaker 2: Tesla to potentially invest in Xai, which will probably happen, 352 00:17:35,320 --> 00:17:37,560 Speaker 2: allowing Musk to continue to pull leverage from his Tesla 353 00:17:37,600 --> 00:17:41,560 Speaker 2: stock until the company's decaying sales and brand eventually swallow 354 00:17:41,640 --> 00:17:45,080 Speaker 2: him whole. But we're not talking about Elon Musk today. 355 00:17:45,840 --> 00:17:48,880 Speaker 2: We are not. We have to talk about Apple now, 356 00:17:48,920 --> 00:17:51,880 Speaker 2: and honestly, they're the least interesting part of this story. 357 00:17:52,080 --> 00:17:54,800 Speaker 2: Their capital expensures in twenty twenty five are expected to 358 00:17:54,840 --> 00:17:57,399 Speaker 2: also be around eleven billion dollars, and they arguably have 359 00:17:57,480 --> 00:18:00,760 Speaker 2: the weirdest AI story in the mag of Since seven, 360 00:18:01,480 --> 00:18:04,800 Speaker 2: Apple intelligence radicalized millions of people against AI, mostly because 361 00:18:04,840 --> 00:18:08,040 Speaker 2: it fucking sucks. Apple clearly got into AI reluctantly and 362 00:18:08,080 --> 00:18:11,159 Speaker 2: now faces stories about how they feel left behind in 363 00:18:11,200 --> 00:18:14,280 Speaker 2: the AI race, which mostly means that Apple aggressively introduce 364 00:18:14,359 --> 00:18:17,640 Speaker 2: people to the actual features of generative AI by force, 365 00:18:17,720 --> 00:18:19,200 Speaker 2: and it turns out that people don't really want to 366 00:18:19,200 --> 00:18:22,800 Speaker 2: summarize documents, or write emails or make custom emoji, and 367 00:18:22,840 --> 00:18:25,600 Speaker 2: anyone who thinks they would is a fucking alien. In 368 00:18:25,640 --> 00:18:27,560 Speaker 2: any case, Apple hasn't bet the farm on AI in 369 00:18:27,640 --> 00:18:30,080 Speaker 2: so much sit It hasn't spent two hundred billion dollars 370 00:18:30,119 --> 00:18:32,119 Speaker 2: in infrastructure for a product of the limited market that 371 00:18:32,200 --> 00:18:34,920 Speaker 2: only loses money. And again, if you want to give 372 00:18:34,920 --> 00:18:36,760 Speaker 2: me some money, I'm gonna put an ad break here. 373 00:18:37,200 --> 00:18:41,400 Speaker 2: So after this, whatever comes next buy it or don't 374 00:18:41,440 --> 00:18:43,320 Speaker 2: if you don't want to, but really you should. If 375 00:18:43,359 --> 00:18:46,000 Speaker 2: I'm speaking, if you hear my voice in the ad, 376 00:18:46,920 --> 00:18:49,239 Speaker 2: then you should buy it unless you don't want it. 377 00:19:05,000 --> 00:19:06,679 Speaker 2: And we're back now. I'm going to use a new 378 00:19:06,760 --> 00:19:09,520 Speaker 2: term I came up with that's really really bad. But 379 00:19:09,880 --> 00:19:13,240 Speaker 2: the fragile five I call them Amazon, Google, Microsoft, Meta, 380 00:19:13,280 --> 00:19:15,840 Speaker 2: and Tesla, the ones investing all the money in the GPUs, 381 00:19:16,000 --> 00:19:18,000 Speaker 2: are holding up the US stock market by funding in 382 00:19:18,080 --> 00:19:21,399 Speaker 2: Vidia's future growth story. And this is really the first 383 00:19:21,400 --> 00:19:23,960 Speaker 2: big takeaway I want you to take from this three parter. 384 00:19:24,480 --> 00:19:25,920 Speaker 2: To be clear, I'm not saying that any of the 385 00:19:25,960 --> 00:19:28,159 Speaker 2: mag seven are going to die, just that five companies 386 00:19:28,160 --> 00:19:30,680 Speaker 2: spend on in Vidio GPUs largely dictate our state of 387 00:19:30,720 --> 00:19:33,720 Speaker 2: the US stock market will be If any of these companies, 388 00:19:33,760 --> 00:19:35,800 Speaker 2: but especially in Vidia, sneeze, you four O one K 389 00:19:35,920 --> 00:19:38,160 Speaker 2: and your kid's college fund will probably catch a cold. 390 00:19:39,359 --> 00:19:42,640 Speaker 2: I realized this sounds a little simplistic, but by my calculations, 391 00:19:42,840 --> 00:19:45,240 Speaker 2: in Vidia's value underpins about eight percent of the value 392 00:19:45,280 --> 00:19:47,520 Speaker 2: of the US stock market. At the time of writing, 393 00:19:47,680 --> 00:19:49,760 Speaker 2: it accounts for roughly seven point five percent of the 394 00:19:49,800 --> 00:19:51,840 Speaker 2: S and P five hundred an index of the five 395 00:19:51,920 --> 00:19:56,080 Speaker 2: hundred largest US publicly traded companies are disturbing eighty eight percent. 396 00:19:56,119 --> 00:19:58,560 Speaker 2: As I mentioned, of in Vidia's revenue comes from enterprise 397 00:19:58,600 --> 00:20:01,960 Speaker 2: scale GPUs, primarily US for Generative AI, of which five 398 00:20:02,000 --> 00:20:05,400 Speaker 2: companies spend, makes up over forty two percent of its revenue. 399 00:20:05,520 --> 00:20:07,560 Speaker 2: In the event that any one of these companies makes 400 00:20:07,560 --> 00:20:10,639 Speaker 2: significant changes to their investments and in video chips, it 401 00:20:10,640 --> 00:20:13,320 Speaker 2: will likely have a direct and meaningful negative impact on 402 00:20:13,359 --> 00:20:17,240 Speaker 2: the wider economy and markets. In Vidia's earnings are effectively 403 00:20:17,280 --> 00:20:20,480 Speaker 2: the US stock market's confidence and everything rides on five companies. 404 00:20:20,720 --> 00:20:23,280 Speaker 2: And if we're honest, here really four companies. This tesla 405 00:20:23,359 --> 00:20:26,960 Speaker 2: is point nine percent of the investment in GPUs of 406 00:20:27,000 --> 00:20:30,880 Speaker 2: those five companies buying GPUs for Generative AI to train 407 00:20:30,960 --> 00:20:34,480 Speaker 2: at generative AI models were still these services while losing. 408 00:20:34,520 --> 00:20:38,480 Speaker 2: These companies' massive amounts of money don't really produce much revenue, 409 00:20:38,520 --> 00:20:40,920 Speaker 2: meaning that the AI trade is not driven by any 410 00:20:40,960 --> 00:20:42,639 Speaker 2: real meaningful revenue growth. 411 00:20:43,080 --> 00:20:46,240 Speaker 3: But d it hid they said? 412 00:20:46,480 --> 00:20:51,080 Speaker 2: They said, points of growth? Silence, quiet, nothing more out 413 00:20:51,160 --> 00:20:54,120 Speaker 2: of you any of these companies talking about growth from 414 00:20:54,119 --> 00:20:56,200 Speaker 2: AI or the jobs that AI will replace, or how 415 00:20:56,280 --> 00:20:59,639 Speaker 2: AI has changed their organization are handwaiving to avoid telling 416 00:20:59,680 --> 00:21:02,440 Speaker 2: you how how much money these services are actually making them. 417 00:21:02,680 --> 00:21:05,040 Speaker 2: If they were making good money and experiencing real growth 418 00:21:05,040 --> 00:21:07,200 Speaker 2: as a result of AI, they wouldn't shut the fuck 419 00:21:07,280 --> 00:21:08,879 Speaker 2: up about it. They'd be in your ear and up 420 00:21:08,880 --> 00:21:10,879 Speaker 2: your ass, hooting and hollering about how much cash they 421 00:21:10,920 --> 00:21:13,359 Speaker 2: were rolling in. And they're not because they're not rolling 422 00:21:13,359 --> 00:21:15,320 Speaker 2: in cash and are in fact blowing nearly one hundred 423 00:21:15,320 --> 00:21:17,960 Speaker 2: billion dollars each to build massive, power hungry, costly data 424 00:21:18,000 --> 00:21:21,720 Speaker 2: centers for no real reason. Don't watch the mouth, watch 425 00:21:21,840 --> 00:21:24,600 Speaker 2: the hands. These companies are going to say they're seeing 426 00:21:24,600 --> 00:21:26,760 Speaker 2: growth from AI, but unless they actually show you the 427 00:21:26,760 --> 00:21:30,240 Speaker 2: growth and enumerate it, they are kind of lying. They're 428 00:21:30,280 --> 00:21:31,600 Speaker 2: lying in the way that you're allowed to. 429 00:21:32,000 --> 00:21:35,360 Speaker 3: But it Hey, Amazon Web Services took years to become profitable. 430 00:21:35,359 --> 00:21:38,159 Speaker 3: People said Amazon would fail. So this is one of 431 00:21:38,160 --> 00:21:41,280 Speaker 3: the most annoying and consistent responses to my work, and 432 00:21:41,320 --> 00:21:43,800 Speaker 3: it's when people say that either Amazon or Amazon Web 433 00:21:43,800 --> 00:21:46,280 Speaker 3: Services ran at a loss. In the Amazon Web Services, 434 00:21:46,320 --> 00:21:49,159 Speaker 3: which pretty much was the invention of modern mass market 435 00:21:49,200 --> 00:21:53,240 Speaker 3: cloud compute infrastructure for running stuff on the cloud, lost money. 436 00:21:52,960 --> 00:21:56,560 Speaker 2: And then didn't. Here's the thing, this statement is one 437 00:21:56,560 --> 00:21:58,880 Speaker 2: of the things that people say because it sounds rational. 438 00:21:58,960 --> 00:22:02,680 Speaker 2: Amazon did lose mone and Amazon Web Services was expensive. 439 00:22:02,720 --> 00:22:05,880 Speaker 2: That's right, right, it's obvious. Right. The thing is, I've 440 00:22:05,880 --> 00:22:07,840 Speaker 2: never really had anyone explain this point to me, so 441 00:22:07,960 --> 00:22:10,240 Speaker 2: I finally sat down. I'm going to deal with this 442 00:22:10,320 --> 00:22:12,919 Speaker 2: criticism because every fucking person who manages it thinks they 443 00:22:13,000 --> 00:22:15,560 Speaker 2: just bulled Excaliper from the stone and can now decapitate me. 444 00:22:16,320 --> 00:22:19,000 Speaker 2: They claim that because people in the past doubted Amazon 445 00:22:19,040 --> 00:22:20,760 Speaker 2: because are in addition to the burn rate of the 446 00:22:20,760 --> 00:22:24,639 Speaker 2: AWS systems as the company built out its infrastructure, that 447 00:22:24,760 --> 00:22:27,719 Speaker 2: I too, am wrong because the analysts were wrong about that. 448 00:22:28,000 --> 00:22:31,320 Speaker 2: This isn't Camelot, You're a rube. You are not King Arthur. 449 00:22:32,160 --> 00:22:34,399 Speaker 2: And now I will address both the argument itself and 450 00:22:34,440 --> 00:22:36,560 Speaker 2: the day part of it too, because if the argument 451 00:22:36,600 --> 00:22:38,880 Speaker 2: is that the people who got Amazon Web Services wrong 452 00:22:38,880 --> 00:22:41,000 Speaker 2: should not be trusted and we should no longer trust them, 453 00:22:41,240 --> 00:22:45,600 Speaker 2: the people who actively propagate and dies something wrong, we 454 00:22:45,640 --> 00:22:50,320 Speaker 2: shouldn't trust them, right right, Well, you'll never guess who's 455 00:22:50,520 --> 00:22:53,240 Speaker 2: now saying AI is good. Oh, I'm going to get there, 456 00:22:53,280 --> 00:22:56,640 Speaker 2: don't you flip and worry? Well, if I'm honest, I'm 457 00:22:56,640 --> 00:22:59,159 Speaker 2: not sure where this argument came from, because there is, 458 00:22:59,200 --> 00:23:01,720 Speaker 2: to my knowledge, no story about Amazon Web Services where 459 00:23:01,760 --> 00:23:05,640 Speaker 2: somebody suggests its burn rate would kill Amazon. But I'm 460 00:23:05,680 --> 00:23:09,360 Speaker 2: a curious little creator, so let's start with an obvious one, 461 00:23:09,400 --> 00:23:11,119 Speaker 2: the obvious point. I want to give a shout out 462 00:23:11,160 --> 00:23:13,000 Speaker 2: to Harry mccrack and a fast company for bringing this 463 00:23:13,000 --> 00:23:15,880 Speaker 2: one up to me. It May May thirty first, ninety 464 00:23:15,960 --> 00:23:17,960 Speaker 2: ninety nine, there was a piece that everybody is thinking 465 00:23:18,000 --> 00:23:21,280 Speaker 2: of called Amazon dot bom and the writer Jacqueline Doherty 466 00:23:21,400 --> 00:23:23,960 Speaker 2: was mocked soundly for being wrong about Amazon, which has 467 00:23:24,040 --> 00:23:26,400 Speaker 2: now become quite profitable. The article, along with the other 468 00:23:26,480 --> 00:23:28,679 Speaker 2: sources that form the basis of this episode, are going 469 00:23:28,680 --> 00:23:31,560 Speaker 2: to be linked in the spreadsheet, and as a as 470 00:23:31,600 --> 00:23:34,159 Speaker 2: a surprise, I'll actually up later. I also want to 471 00:23:34,200 --> 00:23:36,480 Speaker 2: be clear that Amazon Web Services did not launch until 472 00:23:36,480 --> 00:23:39,360 Speaker 2: two thousand and six, and Amazon itself would become reliably 473 00:23:39,359 --> 00:23:43,280 Speaker 2: profitable in two thousand and three. Technically, Amazon had opened 474 00:23:43,359 --> 00:23:46,600 Speaker 2: up Amazon dot COM's web services for developers to incorporate 475 00:23:46,600 --> 00:23:49,200 Speaker 2: Amazon content into their applications in two thousand and two, 476 00:23:49,240 --> 00:23:52,520 Speaker 2: but what we consider Amazon web Services today, cloud storage 477 00:23:52,560 --> 00:23:55,680 Speaker 2: and compute, launched in two thousand and six. But okay, 478 00:23:56,160 --> 00:23:58,960 Speaker 2: fancy pans, what did she actually say? We quote Doherty. 479 00:23:59,480 --> 00:24:02,040 Speaker 2: Unfortunately for Bezos, Amazon is now entering a stage in 480 00:24:02,040 --> 00:24:04,040 Speaker 2: which investors will be less willing to rely on its 481 00:24:04,080 --> 00:24:06,760 Speaker 2: charisma and more demand advances to tough questions like when 482 00:24:06,800 --> 00:24:09,080 Speaker 2: will this company actually turn a profit? And how will 483 00:24:09,119 --> 00:24:11,200 Speaker 2: Amazon triumph over a slew of new competitors who have 484 00:24:11,280 --> 00:24:14,000 Speaker 2: deep pockets and new technologies. Who tried to ask Bezerz, 485 00:24:14,040 --> 00:24:16,120 Speaker 2: but he declined to make himself for any other executives 486 00:24:16,160 --> 00:24:18,840 Speaker 2: of the company available. He could ignore barons, but he 487 00:24:18,840 --> 00:24:21,840 Speaker 2: can't ignore the questions. Bang a line, by the way, 488 00:24:22,200 --> 00:24:24,600 Speaker 2: Amazon last year posted a loss of one hundred and 489 00:24:24,600 --> 00:24:26,560 Speaker 2: twenty five million dollars, which is about two hundred and 490 00:24:26,600 --> 00:24:29,520 Speaker 2: forty two point six million today's money on revenues of 491 00:24:29,560 --> 00:24:31,560 Speaker 2: six hundred and ten millions to about one point one 492 00:24:31,640 --> 00:24:34,359 Speaker 2: eight three billion dollars in today's money, and then this 493 00:24:34,760 --> 00:24:37,760 Speaker 2: year's first quarter referring of course, to nineteen ninety nine, 494 00:24:37,880 --> 00:24:39,760 Speaker 2: as the company posted a loss of sixty one point 495 00:24:39,800 --> 00:24:41,720 Speaker 2: seven million, which is one hundred and nineteen point seventy 496 00:24:41,720 --> 00:24:44,000 Speaker 2: five million today's money on revenues of two hundred and 497 00:24:44,040 --> 00:24:46,240 Speaker 2: ninety three point six million, five hundred and sixty nine 498 00:24:46,240 --> 00:24:48,919 Speaker 2: point eight two million dollars in today's money. I realized 499 00:24:48,920 --> 00:24:51,720 Speaker 2: that was a real motherfucker of a quote, but it's necessary. 500 00:24:52,000 --> 00:24:54,040 Speaker 2: Her argument, for the most part is that Amazon was 501 00:24:54,160 --> 00:24:56,239 Speaker 2: burning cash and had a ton of competition from other 502 00:24:56,240 --> 00:24:58,679 Speaker 2: people doing similar things, and that analysts backed her up, 503 00:24:58,680 --> 00:25:01,440 Speaker 2: and they really did. By the way, again, I quote 504 00:25:01,680 --> 00:25:03,960 Speaker 2: the first mover does not always win the importance of 505 00:25:04,000 --> 00:25:05,840 Speaker 2: being first as a mantra in the internet world, but 506 00:25:05,880 --> 00:25:08,520 Speaker 2: it's wrong. The ones that the most efficient will be successful, 507 00:25:08,560 --> 00:25:11,800 Speaker 2: says one retail analyst. In retailing, anyone can build a 508 00:25:11,800 --> 00:25:13,880 Speaker 2: great looking store. The hard part is building a great 509 00:25:13,920 --> 00:25:16,840 Speaker 2: looking store that makes money, which is a good point 510 00:25:18,000 --> 00:25:21,280 Speaker 2: fair arguments for the time, though perhaps a little narrow minded. 511 00:25:21,320 --> 00:25:23,560 Speaker 2: The assumption wasn't what Amazon was building, and we are, 512 00:25:23,600 --> 00:25:25,760 Speaker 2: by the way, are referring to Amazon dot com. The 513 00:25:25,920 --> 00:25:28,800 Speaker 2: store was a bad idea, but the Amazon wouldn't be 514 00:25:28,840 --> 00:25:31,680 Speaker 2: the ones to build it. And again we quote. Once 515 00:25:31,720 --> 00:25:34,200 Speaker 2: Walmart decides to go after Amazon, there's no contest. The 516 00:25:34,800 --> 00:25:38,960 Speaker 2: Barnard President Bernard's Retail Trend report. Walmart has the resources 517 00:25:38,960 --> 00:25:42,640 Speaker 2: that Amazon can't even dream about, which is true at 518 00:25:42,640 --> 00:25:46,399 Speaker 2: the time, but in simpler terms, Amazon's business model was 519 00:25:46,440 --> 00:25:48,920 Speaker 2: not in question. People were buying shit online. In fact, 520 00:25:49,000 --> 00:25:51,080 Speaker 2: this was just before the dot com bubble burst, when 521 00:25:51,080 --> 00:25:53,440 Speaker 2: people had insane optimism about the future of the web. 522 00:25:53,760 --> 00:25:56,720 Speaker 2: Yet the comparison stops there. People obviously like buying shit online. 523 00:25:56,760 --> 00:25:58,480 Speaker 2: It was the business models of many of these web 524 00:25:58,560 --> 00:26:01,879 Speaker 2: pioneers that sucked at Year WEBVAM. But we're going to 525 00:26:01,880 --> 00:26:04,679 Speaker 2: talk about Amazon Web Services and the less technical of you. 526 00:26:05,000 --> 00:26:08,439 Speaker 2: I want to explain something. AWS is a really important company. 527 00:26:08,440 --> 00:26:10,720 Speaker 2: I'll kind of get into those details. But people like 528 00:26:10,800 --> 00:26:13,040 Speaker 2: to argue about it and say, well, it lost a 529 00:26:13,040 --> 00:26:16,440 Speaker 2: bunch of money, so you know that means that generative 530 00:26:16,480 --> 00:26:17,840 Speaker 2: a I should lose a bunch of money to and 531 00:26:17,880 --> 00:26:21,080 Speaker 2: that's how it works. I'm going to substantively and repeatedly 532 00:26:21,119 --> 00:26:24,480 Speaker 2: explain why that is so goddamn stupid. I'm sick of 533 00:26:24,480 --> 00:26:29,359 Speaker 2: the argument. I'm sick of it. Breathe Edward, breathe. They 534 00:26:29,400 --> 00:26:34,439 Speaker 2: can't get you behind the microphone, okay. Amazon Web Services 535 00:26:34,600 --> 00:26:37,119 Speaker 2: was an outgrowth of Amazon's own infrastructure, which had to 536 00:26:37,160 --> 00:26:39,439 Speaker 2: expand rapidly to deal with the influx of web traffic 537 00:26:39,440 --> 00:26:41,399 Speaker 2: from Amazon Dot Com, which had become one of the 538 00:26:41,400 --> 00:26:44,320 Speaker 2: world's most popular websites and was becoming increasingly more complex 539 00:26:44,359 --> 00:26:47,240 Speaker 2: as it sold things other than books to multiple international 540 00:26:47,280 --> 00:26:50,680 Speaker 2: locations as well. Other companies had created their own infrastructure, 541 00:26:50,680 --> 00:26:52,719 Speaker 2: but if a smaller company wanted to scale, they basically 542 00:26:52,800 --> 00:26:54,960 Speaker 2: needed to build their own thing. It was a massive 543 00:26:55,040 --> 00:26:58,760 Speaker 2: barrier between companies and building web services. And it's actually 544 00:26:58,920 --> 00:27:01,360 Speaker 2: kind of cool what Amazon did. I hate to look 545 00:27:01,440 --> 00:27:04,359 Speaker 2: Rosie I bind rose colored lenses. I don't know the 546 00:27:04,400 --> 00:27:08,080 Speaker 2: phrase that Jeff Bezos, but oh no, I remember this 547 00:27:08,160 --> 00:27:10,719 Speaker 2: was early two thousands, before Facebook, Twitter, and a lot 548 00:27:10,760 --> 00:27:14,000 Speaker 2: of modern internet we know that runs on services like 549 00:27:14,040 --> 00:27:17,359 Speaker 2: Amazon Web Services or Microsoft Zero or Google Cloud. They 550 00:27:17,400 --> 00:27:19,879 Speaker 2: basically invented the modern concept of cloud compume. 551 00:27:20,359 --> 00:27:21,280 Speaker 3: But we're bit to. 552 00:27:21,240 --> 00:27:24,639 Speaker 2: Talk about Amazon Web Services being dangerous for Amazon and 553 00:27:24,680 --> 00:27:28,000 Speaker 2: people hating on it the thing that allegedly happened, right. 554 00:27:28,920 --> 00:27:30,760 Speaker 2: I do hope all the people that said this to 555 00:27:30,800 --> 00:27:33,240 Speaker 2: me didn't just make it up. Oh my god, they did. 556 00:27:33,600 --> 00:27:36,040 Speaker 2: A November two thousand and six story from Bloomberg talked 557 00:27:36,040 --> 00:27:39,080 Speaker 2: about Jeff Bezos is risky bet to run your business 558 00:27:39,080 --> 00:27:41,880 Speaker 2: with the technology behind his website, saying that that wall 559 00:27:41,920 --> 00:27:44,840 Speaker 2: Street wanted him to mind the store. Bezos referred to 560 00:27:44,880 --> 00:27:46,879 Speaker 2: as a one time Internet poster boy that became a 561 00:27:46,960 --> 00:27:50,280 Speaker 2: post dot com pinata. Fuck, they were so good, but 562 00:27:50,600 --> 00:27:54,000 Speaker 2: where is this pisson vinegar? By the way, this is fun. Nevertheless, 563 00:27:54,080 --> 00:27:56,359 Speaker 2: this article, which again is linked in the spreadsheet for 564 00:27:56,359 --> 00:27:59,480 Speaker 2: the episode notes, has what I think my hater's crave 565 00:28:00,080 --> 00:28:02,840 Speaker 2: hmm and I quote. But if techies are wowed by 566 00:28:02,880 --> 00:28:05,679 Speaker 2: Bezoz's grand plan, it's not likely to win many converts 567 00:28:05,680 --> 00:28:08,399 Speaker 2: on Wall Street. To many observers, it conjures up the 568 00:28:08,400 --> 00:28:11,240 Speaker 2: ghost of Amazon past. During the dot com boom, Bezez 569 00:28:11,240 --> 00:28:13,720 Speaker 2: spent hundreds of millions of dollars to build distribution centers 570 00:28:13,720 --> 00:28:15,680 Speaker 2: and computer systems in the promise that they would eventually 571 00:28:15,720 --> 00:28:18,160 Speaker 2: pay off with outsize returns. That helps set the stage 572 00:28:18,160 --> 00:28:20,760 Speaker 2: of the world's biggest web retail operation, with expected sales 573 00:28:20,800 --> 00:28:24,880 Speaker 2: at ten point five billion dollars this year. All that 574 00:28:24,960 --> 00:28:27,840 Speaker 2: has investors, wrestlers, and many analysts throwing up their hands 575 00:28:27,880 --> 00:28:30,520 Speaker 2: wondering if Bezos is merely flailing around from an alternative 576 00:28:30,560 --> 00:28:33,639 Speaker 2: to his retail operation. Eleven of twenty seven analysts who 577 00:28:33,640 --> 00:28:36,320 Speaker 2: followed the company of under perform or sell ratings on 578 00:28:36,359 --> 00:28:39,680 Speaker 2: the stock a stunning vote of no confidence. That number 579 00:28:39,680 --> 00:28:43,000 Speaker 2: of cell recommendations is matched among large companies only by 580 00:28:43,080 --> 00:28:47,400 Speaker 2: Quest Communications International, Inc. According to investment consultants Starmind Corp. 581 00:28:47,880 --> 00:28:49,920 Speaker 2: It's more than even the eight cell options on the 582 00:28:49,960 --> 00:28:54,320 Speaker 2: struggling Ford Motor Company. Pretty bad, right, pretty bad. My 583 00:28:54,360 --> 00:28:57,400 Speaker 2: goose is cooked, all those analysts seem pretty bad, except 584 00:28:57,480 --> 00:28:59,880 Speaker 2: it's not. My goose is raw. Yours, however, has been 585 00:28:59,920 --> 00:29:01,880 Speaker 2: in the oven for over a year. As one on 586 00:29:02,040 --> 00:29:05,000 Speaker 2: as Scott W. Duvitt noted at the time, the direct 587 00:29:05,000 --> 00:29:09,840 Speaker 2: costs of providing Amazon Web services at first were miniscule 588 00:29:10,360 --> 00:29:13,600 Speaker 2: because much of the startup infrastructure already existed. It was 589 00:29:13,640 --> 00:29:18,040 Speaker 2: sur plus capacity, Amazon already owned, software Amazon already used, 590 00:29:18,040 --> 00:29:20,160 Speaker 2: and Amazon was in it for the long haul. It 591 00:29:20,240 --> 00:29:22,360 Speaker 2: knew that this would take some time before it became 592 00:29:22,400 --> 00:29:24,920 Speaker 2: a profitable business unit, as the company was basically scaling 593 00:29:24,960 --> 00:29:28,720 Speaker 2: up the infrastructure of the Internet. And by the way, 594 00:29:28,960 --> 00:29:31,600 Speaker 2: let's just go back to that quote here. The quote 595 00:29:31,680 --> 00:29:35,760 Speaker 2: says the costs were miniscule. The costs weren't the problem. 596 00:29:35,800 --> 00:29:39,360 Speaker 2: Hey wait a second, that's a name, Scott W. Duvit. 597 00:29:39,680 --> 00:29:41,440 Speaker 2: I can look him up. I wonder what he's up 598 00:29:41,480 --> 00:29:44,360 Speaker 2: to right now? Oh oh, looks like he's working at 599 00:29:44,400 --> 00:29:47,640 Speaker 2: web Bush as it's managing director of equity. Researcher has 600 00:29:47,720 --> 00:29:50,040 Speaker 2: said that AI companies would enter a new stage and 601 00:29:50,440 --> 00:29:53,720 Speaker 2: early twenty five he said, Oh oh God, just listen 602 00:29:53,760 --> 00:29:56,960 Speaker 2: to this. The second stage is the application phase of 603 00:29:57,000 --> 00:29:59,360 Speaker 2: the cycle, which would benefit software companies as well as 604 00:29:59,360 --> 00:30:01,640 Speaker 2: the cloud of I. And then phase three of this 605 00:30:01,760 --> 00:30:04,440 Speaker 2: will ultimately be the consumer facing companies figuring out how 606 00:30:04,440 --> 00:30:06,520 Speaker 2: to use the technology in ways that can actually drive 607 00:30:06,720 --> 00:30:10,040 Speaker 2: increased interactions with customers. The analyst says the market will 608 00:30:10,120 --> 00:30:12,640 Speaker 2: enter phase two in twenty twenty five, with software companies 609 00:30:12,640 --> 00:30:15,400 Speaker 2: and collaborvider stocks expected to see gains. He adds that 610 00:30:15,440 --> 00:30:20,640 Speaker 2: cybersecurity companies could also benefits the technology evolves. I know 611 00:30:20,720 --> 00:30:24,040 Speaker 2: I'm meant to be more mature, but it also calls 612 00:30:24,040 --> 00:30:27,840 Speaker 2: out Palenteer, Snowflake, and salesforces those who would gain. In 613 00:30:27,840 --> 00:30:29,440 Speaker 2: none of these cases am I able to see any 614 00:30:29,480 --> 00:30:32,720 Speaker 2: actual revenue from AI. Salesforce themselves cerid according to the 615 00:30:32,720 --> 00:30:35,080 Speaker 2: information that they'd see no revenue growth from AI in 616 00:30:35,120 --> 00:30:39,080 Speaker 2: twenty twenty five. Palenteer also has discovered by the Autonomy 617 00:30:39,080 --> 00:30:42,520 Speaker 2: Institute's recent study recently added the following to its public disclosures. 618 00:30:43,600 --> 00:30:46,200 Speaker 2: There are significant risks involved in deploying AI, and there 619 00:30:46,200 --> 00:30:48,400 Speaker 2: could be no assurance that using AI in our platforms 620 00:30:48,400 --> 00:30:51,320 Speaker 2: and products will enhance be beneficial to our business, including 621 00:30:51,320 --> 00:30:54,200 Speaker 2: our profitability. What I'm trying to say here is that 622 00:30:54,200 --> 00:30:56,240 Speaker 2: analysts can be wrong, and they could be wrong at scale. 623 00:30:56,280 --> 00:30:59,240 Speaker 2: There is no analyst consensus that agrees with me. In fact, 624 00:30:59,240 --> 00:31:02,040 Speaker 2: most analysts to be bullish and AI despite the significantly 625 00:31:02,080 --> 00:31:04,920 Speaker 2: worse costs and lack total lack of growth, but ed 626 00:31:05,240 --> 00:31:08,320 Speaker 2: ed Amazon Web Services cost money ahead. Now you should 627 00:31:08,320 --> 00:31:11,160 Speaker 2: meet your awn nice S tried chuckles. In twenty fifteen, 628 00:31:11,200 --> 00:31:14,200 Speaker 2: the year that Amazon Web Services became profitable, Morgan Stanley 629 00:31:14,240 --> 00:31:17,000 Speaker 2: analyst Catty Huberty believed that it was running and a 630 00:31:17,040 --> 00:31:19,960 Speaker 2: material loss, suggesting that the five point five billion dollars 631 00:31:20,000 --> 00:31:24,400 Speaker 2: of Amazon's technology and content expenses was actually AWS expenses 632 00:31:24,600 --> 00:31:27,040 Speaker 2: with a negative contribution of one point three billion dollars. 633 00:31:27,680 --> 00:31:30,080 Speaker 2: And by the way, want to know what she's up 634 00:31:30,080 --> 00:31:33,960 Speaker 2: to nowadays? I wanted to know because six months ago 635 00:31:34,080 --> 00:31:36,160 Speaker 2: she declared that twenty twenty five would be the year 636 00:31:36,200 --> 00:31:40,120 Speaker 2: of AGENTICAI robust they enterprise adoption and broadening AI winners. 637 00:31:40,800 --> 00:31:43,960 Speaker 2: So yes, analysts really got AWS wrong. But putting that aside, 638 00:31:43,960 --> 00:31:47,000 Speaker 2: there might actually be a comparison here. Amazon Web Services 639 00:31:47,080 --> 00:31:51,000 Speaker 2: absolutely created the capital expenditures drain on Amazon from Forbes's 640 00:31:51,040 --> 00:31:54,440 Speaker 2: Chuck Jones. In twenty fourteen, Amazon at four point nine 641 00:31:54,480 --> 00:31:57,000 Speaker 2: billion dollars in capital expenditures, up forty two percent from 642 00:31:57,000 --> 00:31:59,760 Speaker 2: twenty thirteen to three point four billion dollars. The company 643 00:31:59,800 --> 00:32:01,520 Speaker 2: is a wide range of items that it buys the 644 00:32:01,520 --> 00:32:04,480 Speaker 2: support and grow its businesses, ranging from warehouses, robots, and 645 00:32:04,480 --> 00:32:08,000 Speaker 2: computer systems for its core retail business and AAWS. Well, 646 00:32:08,000 --> 00:32:11,520 Speaker 2: I don't expect Amazon to detail how much goes to AWS. 647 00:32:11,560 --> 00:32:14,240 Speaker 2: I suspect it is a decent percentage, which means Amazon 648 00:32:14,880 --> 00:32:18,960 Speaker 2: needs to generate appropriate returns on the capitol deployed from AWS. 649 00:32:19,400 --> 00:32:21,720 Speaker 2: In today's money, this means that Amazon spent six point 650 00:32:21,760 --> 00:32:25,440 Speaker 2: seven billion dollars in capital expenditures in twenty fourteen, likely 651 00:32:25,480 --> 00:32:28,920 Speaker 2: on AWS, assuming it was this much every year. It wasn't, 652 00:32:28,960 --> 00:32:30,640 Speaker 2: but I want to make an example of every person 653 00:32:30,680 --> 00:32:33,000 Speaker 2: claiming that this is a gotcha. It took sixty seven 654 00:32:33,080 --> 00:32:35,880 Speaker 2: point six billion dollars, and that's in today's money, and 655 00:32:36,080 --> 00:32:39,520 Speaker 2: about nine or ten years of pure capital expenditures, even 656 00:32:39,560 --> 00:32:42,800 Speaker 2: though all that capex wasn't just AWS to turn Amazon 657 00:32:42,840 --> 00:32:45,360 Speaker 2: Web Services into a business that now makes billions of 658 00:32:45,440 --> 00:32:49,320 Speaker 2: dollars a quarter in profit. And that's fifteen point four 659 00:32:49,400 --> 00:32:52,840 Speaker 2: billion dollars less than Amazon's capex for twenty twenty four, 660 00:32:53,720 --> 00:32:56,680 Speaker 2: and even less than one hundred and five billion dollars 661 00:32:56,720 --> 00:32:59,360 Speaker 2: they spent this year. It's a fucking joke. And to 662 00:32:59,400 --> 00:33:02,400 Speaker 2: be clear, the actual capital expension numbers are like the 663 00:33:02,520 --> 00:33:06,720 Speaker 2: AWS cost in totality, we're likely much lower. I just 664 00:33:06,760 --> 00:33:09,160 Speaker 2: want to make it clear that even when factoring in inflation, 665 00:33:09,280 --> 00:33:12,320 Speaker 2: AWS was a a bargain and b a fraction of 666 00:33:12,360 --> 00:33:14,560 Speaker 2: the cost of what Amazon is spent in twenty twenty 667 00:33:14,560 --> 00:33:18,320 Speaker 2: four or twenty twenty five. Here's a funny little thing. 668 00:33:18,440 --> 00:33:21,480 Speaker 2: On March thirtieth, twenty fifteen, New York Magazine published a 669 00:33:21,480 --> 00:33:24,160 Speaker 2: piece from none other than mister Kevin Ruse about the 670 00:33:24,160 --> 00:33:27,000 Speaker 2: cloud compute wars, in which he claimed that, and I quote, 671 00:33:27,640 --> 00:33:29,640 Speaker 2: there's no reason to suspect that Amazon would ever need 672 00:33:29,680 --> 00:33:32,120 Speaker 2: to raise prices in AWS or turn the fabled profits 673 00:33:32,120 --> 00:33:35,080 Speaker 2: which the pundits have been speculating about for years less 674 00:33:35,080 --> 00:33:37,760 Speaker 2: than a month later, Amazon revealed that Amazon Web Services 675 00:33:37,800 --> 00:33:40,280 Speaker 2: was profitable. They don't call him the most right man 676 00:33:40,320 --> 00:33:42,840 Speaker 2: in tech journalism for nothing. I think it's so funny 677 00:33:42,880 --> 00:33:45,000 Speaker 2: when you go back and read all of Kevin Rus's stuff, 678 00:33:45,160 --> 00:33:47,880 Speaker 2: how many just like rates he steps on, and how 679 00:33:48,440 --> 00:33:50,560 Speaker 2: quickly they whammy him in the face, and how no 680 00:33:51,520 --> 00:33:55,800 Speaker 2: body says anything. I'm saying something. But here's the good news. 681 00:33:55,840 --> 00:33:58,200 Speaker 2: We're at the end of this first part. Next episode, 682 00:33:58,200 --> 00:34:00,880 Speaker 2: we're going to continue exploring the comparison in AWS and 683 00:34:00,920 --> 00:34:03,719 Speaker 2: general of AI and talk about why that comparison fundamentally 684 00:34:03,720 --> 00:34:07,080 Speaker 2: doesn't work and why everything's cry the bripple. I'll catch 685 00:34:07,120 --> 00:34:08,200 Speaker 2: you on the flip side. 686 00:34:08,239 --> 00:34:19,200 Speaker 1: Thanks for listening, Thank you for listening to Better Offline. 687 00:34:19,320 --> 00:34:21,719 Speaker 4: The editor and composer of the Better Offline theme song 688 00:34:21,800 --> 00:34:24,439 Speaker 4: is Metasowski. You can check out more of his music 689 00:34:24,480 --> 00:34:28,120 Speaker 4: and audio projects at Matasowski dot com, M A T 690 00:34:28,120 --> 00:34:32,600 Speaker 4: T O, S O W s ki dot com. You 691 00:34:32,600 --> 00:34:35,120 Speaker 4: can email me at easy at Better Offline dot com, 692 00:34:35,239 --> 00:34:37,560 Speaker 4: or visit Better Offline dot com to find more podcast 693 00:34:37,600 --> 00:34:40,920 Speaker 4: links and of course, my newsletter. I also really recommend 694 00:34:40,960 --> 00:34:42,880 Speaker 4: you go to chat dot Where's Youreed dot at to 695 00:34:42,960 --> 00:34:45,560 Speaker 4: visit the discord and go to our slash Better. 696 00:34:45,280 --> 00:34:48,480 Speaker 1: Offline to check out our reddit. Thank you so much 697 00:34:48,480 --> 00:34:52,320 Speaker 1: for listening. Better Offline is a production of cool Zone Media. 698 00:34:52,480 --> 00:34:55,319 Speaker 2: For more from cool Zone Media, visit our website cool 699 00:34:55,360 --> 00:34:58,719 Speaker 2: zonemedia dot com, or check us out on the iHeartRadio app, 700 00:34:58,760 --> 00:35:14,680 Speaker 2: Apple Podcasts, whever you get your podcasts.