1 00:00:05,680 --> 00:00:11,840 Speaker 1: Microsoft, Word, power Point, Excel. They are fixtures of office 2 00:00:11,840 --> 00:00:16,080 Speaker 1: life everywhere, but it's almost thirty years since Microsoft released 3 00:00:16,079 --> 00:00:20,079 Speaker 1: the iconic Office ninety five. In the space of just 4 00:00:20,160 --> 00:00:23,160 Speaker 1: a few months, though, the company that once gave us 5 00:00:23,200 --> 00:00:27,240 Speaker 1: the likes of Clippy the interactive paper clip, has suddenly 6 00:00:27,280 --> 00:00:30,560 Speaker 1: become one of the most important players in the rapid 7 00:00:30,640 --> 00:00:34,479 Speaker 1: rise of artificial intelligence, and it all has to do 8 00:00:34,560 --> 00:00:38,320 Speaker 1: with its collaboration with open AI, the creator of the 9 00:00:38,440 --> 00:00:42,200 Speaker 1: wildly popular chat GPT that seems to be taking over 10 00:00:42,240 --> 00:00:44,760 Speaker 1: the world and that some people fear actually will take 11 00:00:44,800 --> 00:00:45,720 Speaker 1: over the world. 12 00:00:46,080 --> 00:00:47,800 Speaker 2: If you went back a couple of years and you 13 00:00:47,880 --> 00:00:50,240 Speaker 2: asked people at Silicon Valley, like, who's in the lead 14 00:00:50,400 --> 00:00:53,560 Speaker 2: in AI, they would have said Google. And I think 15 00:00:53,640 --> 00:00:57,200 Speaker 2: everyone agrees now that Microsoft is either kind of like 16 00:00:57,320 --> 00:00:59,880 Speaker 2: in the lead or has at least gotten to parody 17 00:01:00,000 --> 00:01:00,480 Speaker 2: with Google. 18 00:01:06,880 --> 00:01:10,080 Speaker 1: I'm West Kasova today on the Big Take. Bloomberg BusinessWeek 19 00:01:10,120 --> 00:01:13,520 Speaker 1: writer Max Chafkin is here with the unlikely story of 20 00:01:13,560 --> 00:01:21,399 Speaker 1: how Microsoft, of all companies, is making itself cool again. Max, 21 00:01:21,480 --> 00:01:25,319 Speaker 1: We've obviously been talking a lot lately about open AI 22 00:01:25,640 --> 00:01:30,520 Speaker 1: and chat gpt, all the different versions of AI. But 23 00:01:30,680 --> 00:01:34,520 Speaker 1: how does Microsoft enter into this because they are really 24 00:01:34,560 --> 00:01:36,959 Speaker 1: a central player in this whole story. 25 00:01:38,280 --> 00:01:41,400 Speaker 2: Yeah. Absolutely, when you think about these kind of crazy 26 00:01:41,440 --> 00:01:45,200 Speaker 2: developments of the last i don't know, eight nine months, 27 00:01:45,240 --> 00:01:48,360 Speaker 2: the release of chat gpt and then this kind of 28 00:01:48,440 --> 00:01:52,280 Speaker 2: frenzy of activity related ai. You know, one player, as 29 00:01:52,280 --> 00:01:54,720 Speaker 2: you said, one of the key players is open ai, 30 00:01:54,800 --> 00:01:58,920 Speaker 2: which is the startup that created chat gpt. But basically 31 00:01:59,000 --> 00:02:03,120 Speaker 2: an equally big player is Microsoft, which is the biggest 32 00:02:03,120 --> 00:02:05,920 Speaker 2: investor in open Ai. But that even kind of understates 33 00:02:05,960 --> 00:02:10,080 Speaker 2: how important Microsoft is to this startup. Microsoft is the 34 00:02:10,160 --> 00:02:12,520 Speaker 2: largest shareholder, It owns something like forty nine percent of 35 00:02:12,560 --> 00:02:16,560 Speaker 2: the company. It is the sole provider of computing services. 36 00:02:16,600 --> 00:02:19,839 Speaker 2: It runs all the computers that open ai uses. It's 37 00:02:19,880 --> 00:02:22,280 Speaker 2: also the main commercial partner. It's the way that open 38 00:02:22,320 --> 00:02:24,800 Speaker 2: ai gets revenue. And you know, as we wrote in 39 00:02:24,800 --> 00:02:28,400 Speaker 2: the story, I think it's probably useful to think about 40 00:02:28,400 --> 00:02:30,799 Speaker 2: open Ai not so much as a startup, but as 41 00:02:30,840 --> 00:02:35,560 Speaker 2: a very successful, you know, quasi independent subsidiary of the 42 00:02:35,680 --> 00:02:38,359 Speaker 2: largest business software company in the world, the second largest 43 00:02:38,360 --> 00:02:40,239 Speaker 2: company in the world by marketcap. 44 00:02:41,240 --> 00:02:43,960 Speaker 1: And how that partnership came to be as the subject 45 00:02:44,000 --> 00:02:46,280 Speaker 1: of this big story you've written, and it's really interesting 46 00:02:46,320 --> 00:02:50,200 Speaker 1: because Microsoft is huge, we all use their products, but 47 00:02:50,280 --> 00:02:53,200 Speaker 1: I don't think people have thought about Microsoft as a 48 00:02:53,280 --> 00:02:56,560 Speaker 1: cutting edge, innovative company for a really long time. 49 00:02:57,840 --> 00:02:59,880 Speaker 2: When my co writer Dina Bass and I were in 50 00:03:00,040 --> 00:03:03,320 Speaker 2: Redmond in Washington near Seattle, on the Microsoft campus, you know, 51 00:03:03,320 --> 00:03:05,760 Speaker 2: they were talking about how amazing it was, and they 52 00:03:05,800 --> 00:03:08,799 Speaker 2: were saying, you know, it's like Windows ninety five all 53 00:03:08,840 --> 00:03:11,520 Speaker 2: over again, at which I loved because, first of all, 54 00:03:11,680 --> 00:03:14,560 Speaker 2: to most people outside of Redmen that is like an insult, 55 00:03:14,880 --> 00:03:17,360 Speaker 2: but if you're there, it's the highest compliment because that 56 00:03:17,480 --> 00:03:20,000 Speaker 2: is really the time when Microsoft was on top of 57 00:03:20,040 --> 00:03:23,000 Speaker 2: the world. And since then, you know, they went through 58 00:03:23,040 --> 00:03:25,760 Speaker 2: a very sort of dark period in the sort of 59 00:03:25,800 --> 00:03:30,680 Speaker 2: two thousands. Ever since Sati Nadella became CEO of Microsoft, 60 00:03:30,840 --> 00:03:33,360 Speaker 2: they've been on this kind of turnaround trajectory. And some 61 00:03:33,440 --> 00:03:36,080 Speaker 2: of that has had to do with business decisions that 62 00:03:36,160 --> 00:03:38,720 Speaker 2: he's made, you know, having to do with basically playing 63 00:03:38,800 --> 00:03:42,000 Speaker 2: nicer with other platforms like Apple, but a big part 64 00:03:42,000 --> 00:03:44,880 Speaker 2: of it is this AI thing and especially the deal 65 00:03:45,000 --> 00:03:49,880 Speaker 2: with open Ai. And really this all started at Sun Valley, 66 00:03:49,920 --> 00:03:53,080 Speaker 2: which is this annual conference organized by Allen and Company. 67 00:03:53,080 --> 00:03:56,720 Speaker 2: It's an investment bank it's in Idaho, and it's essentially 68 00:03:57,000 --> 00:04:00,320 Speaker 2: billionaire summer camp. Billionaires go there to hang out out 69 00:04:00,400 --> 00:04:02,920 Speaker 2: and make deals. It's the one place you see all 70 00:04:02,960 --> 00:04:05,160 Speaker 2: these guys like looking sort of relaxed. If you see 71 00:04:05,160 --> 00:04:07,800 Speaker 2: a picture of a CEO wearing a lanyard, it's almost 72 00:04:07,800 --> 00:04:11,080 Speaker 2: definitely a Sun Valley picture, you know. And Altman, Sam Altman, 73 00:04:11,280 --> 00:04:13,520 Speaker 2: the co founder of open Ai. You know, he's a 74 00:04:13,520 --> 00:04:15,720 Speaker 2: guy who goes to Sun Valley. He is a young guy. 75 00:04:15,760 --> 00:04:18,160 Speaker 2: He's only about forty, but he's sort of been in 76 00:04:18,320 --> 00:04:20,680 Speaker 2: Silicon Valley for a really long time. He started a 77 00:04:20,680 --> 00:04:23,240 Speaker 2: company called Looped, which was like a mobile startup. He 78 00:04:23,360 --> 00:04:26,880 Speaker 2: ran y combinator, you know, big deal kind of investment firm. 79 00:04:27,040 --> 00:04:30,640 Speaker 2: He's sort of everywhere. He's friends with everybody. And at 80 00:04:30,800 --> 00:04:34,000 Speaker 2: Sun Valley, he and Nadella run into each other and 81 00:04:34,040 --> 00:04:37,160 Speaker 2: they get to talking. And as we write in the story, 82 00:04:37,360 --> 00:04:40,560 Speaker 2: you know, a couple of things became clear in this conversation, 83 00:04:40,600 --> 00:04:43,320 Speaker 2: which then continued and culminated in a deal in twenty nineteen, 84 00:04:43,640 --> 00:04:46,440 Speaker 2: one of which is that Sam Altman has sort of 85 00:04:46,480 --> 00:04:51,120 Speaker 2: these really big ideas, sort of famously big. He thinks 86 00:04:51,200 --> 00:04:54,440 Speaker 2: that AI could potentially end the world he thinks that 87 00:04:54,520 --> 00:04:58,920 Speaker 2: he is creating artificial superintelligence that's potentially you know, lead 88 00:04:59,000 --> 00:05:01,320 Speaker 2: us to a life life of leisure where there are 89 00:05:01,320 --> 00:05:03,240 Speaker 2: no jobs, or there are very few jobs, which could 90 00:05:03,279 --> 00:05:06,320 Speaker 2: have huge societal consequences. But the thing I think that 91 00:05:06,360 --> 00:05:09,719 Speaker 2: appealed to Microsoft is that open ai was actually making 92 00:05:09,760 --> 00:05:15,080 Speaker 2: progress towards commercial products. Microsoft had invested essentially huge sums 93 00:05:15,080 --> 00:05:17,120 Speaker 2: in AI, and they'd made a lot of progress. They'd 94 00:05:17,120 --> 00:05:19,320 Speaker 2: written a lot of good research papers, you know, as 95 00:05:19,400 --> 00:05:21,840 Speaker 2: Nadella and the other Microsoft executive saw it, but they 96 00:05:21,960 --> 00:05:24,560 Speaker 2: hadn't really had products, right. It was a lot of 97 00:05:24,680 --> 00:05:27,800 Speaker 2: really good ideas, but kind of disorganized. And in open 98 00:05:27,839 --> 00:05:31,159 Speaker 2: Ai they saw something you know, much much more focus. 99 00:05:31,680 --> 00:05:34,280 Speaker 2: And that was the thing I think that most appealed 100 00:05:34,320 --> 00:05:37,840 Speaker 2: to Microsoft. That's why they invested in twenty nineteen a 101 00:05:37,920 --> 00:05:40,240 Speaker 2: billion dollars in the company. And that's kind of like 102 00:05:40,279 --> 00:05:41,680 Speaker 2: the beginning of this relationship. 103 00:05:43,240 --> 00:05:46,000 Speaker 1: And where did that money go? How did they then 104 00:05:46,320 --> 00:05:49,800 Speaker 1: use it to rapidly expand this platform. 105 00:05:50,760 --> 00:05:53,640 Speaker 2: Well that's what's so interesting because we talked to Sam 106 00:05:53,720 --> 00:05:55,640 Speaker 2: Altman about this and we said, why didn't you go 107 00:05:55,720 --> 00:05:58,200 Speaker 2: with a normal you know, venture capital firm, And he said, 108 00:05:58,200 --> 00:06:00,839 Speaker 2: the reason is essentially that this kind of technology, the 109 00:06:00,880 --> 00:06:04,760 Speaker 2: thing that they were building, right, it needs these gigantic computers, 110 00:06:04,800 --> 00:06:09,960 Speaker 2: these very very specific configurations of servers with very specific 111 00:06:10,040 --> 00:06:13,719 Speaker 2: chips that cost enormous amounts of money. We're talking hundreds 112 00:06:13,760 --> 00:06:17,279 Speaker 2: of millions, if not billions of dollars just to train 113 00:06:17,400 --> 00:06:20,960 Speaker 2: the model, just to create the model that underlies chat GPT. 114 00:06:21,080 --> 00:06:22,920 Speaker 2: And that's even before you start running it, you start 115 00:06:22,920 --> 00:06:26,080 Speaker 2: answering people's questions. And so the Microsoft deal, right it 116 00:06:26,120 --> 00:06:28,720 Speaker 2: gets reported as a billion dollars and since then, as 117 00:06:28,720 --> 00:06:31,039 Speaker 2: we write in the story, Microsoft has put something like 118 00:06:31,080 --> 00:06:34,360 Speaker 2: thirteen billion into the company in total. But a lot 119 00:06:34,400 --> 00:06:37,200 Speaker 2: of that money, I'd say probably most of that money 120 00:06:37,440 --> 00:06:40,080 Speaker 2: goes right back to Microsoft because open Ai has a 121 00:06:40,120 --> 00:06:44,000 Speaker 2: deal with Microsoft where Microsoft is their cloud provider, and 122 00:06:44,040 --> 00:06:47,719 Speaker 2: that's the big cost here. So Nadella writes Sam Altman 123 00:06:47,760 --> 00:06:50,480 Speaker 2: a check. Sam Altman turns around and starts buying these 124 00:06:50,560 --> 00:06:53,000 Speaker 2: you know, basically time in Azu, which is the Microsoft 125 00:06:53,040 --> 00:06:56,720 Speaker 2: cloud computing business. So that kind of works for everybody, right, 126 00:06:56,800 --> 00:07:00,240 Speaker 2: because on one hand, it seems like Microsoft is is 127 00:07:00,480 --> 00:07:03,240 Speaker 2: outsourcing this really important thing. It's sort of turning over 128 00:07:03,320 --> 00:07:07,719 Speaker 2: a really important piece of intellectual property to an outside company. 129 00:07:08,040 --> 00:07:10,840 Speaker 2: On the other hand, what they're doing is they're building 130 00:07:10,840 --> 00:07:13,800 Speaker 2: this Ai supercomputer, that's what they call it, and that 131 00:07:13,920 --> 00:07:17,600 Speaker 2: is Microsoft intellectual property, that's theirs. They're obviously going to 132 00:07:17,640 --> 00:07:19,760 Speaker 2: continue to sell it to open Ai, but they can 133 00:07:19,800 --> 00:07:22,880 Speaker 2: also sell it to other people. And as these large 134 00:07:22,920 --> 00:07:27,320 Speaker 2: language models, these chatchept like services take off, they have 135 00:07:27,400 --> 00:07:29,480 Speaker 2: the potential to make money not just from open Ai, 136 00:07:29,600 --> 00:07:33,400 Speaker 2: not just from open ai users, but potentially from everybody. 137 00:07:33,160 --> 00:07:36,520 Speaker 1: And Max, how does OpenAI feel about this arrangement. 138 00:07:37,040 --> 00:07:39,480 Speaker 2: Yeah, well, we talked to Sam Altman about it and 139 00:07:39,520 --> 00:07:41,440 Speaker 2: he feels good about it. He feels like it's the 140 00:07:41,440 --> 00:07:43,920 Speaker 2: only way that he can do what he wants to do. 141 00:07:44,840 --> 00:07:48,680 Speaker 1: And you write that the critical I guess breakthrough was 142 00:07:48,720 --> 00:07:52,240 Speaker 1: something called transfer learning. Can you describe what that wasn't 143 00:07:52,240 --> 00:07:53,560 Speaker 1: why it was such a big deal. 144 00:07:55,080 --> 00:07:59,200 Speaker 2: Yeah, So, if you think back to twenty nineteen, the 145 00:07:59,240 --> 00:08:01,680 Speaker 2: hot thing was dryverless cars. The way you train a 146 00:08:01,760 --> 00:08:06,040 Speaker 2: driverless car is you basically have a bunch of people 147 00:08:06,120 --> 00:08:08,800 Speaker 2: sitting in rooms and they're looking at pictures of road 148 00:08:08,880 --> 00:08:11,760 Speaker 2: signs and they they're tagging them. They're saying, that's a roadsign, 149 00:08:11,800 --> 00:08:13,960 Speaker 2: that's a stop sign, that's a speed limit sign. You 150 00:08:14,000 --> 00:08:16,200 Speaker 2: feed that into the machine and you try to get 151 00:08:16,200 --> 00:08:19,320 Speaker 2: it to identify the sign. So you're taking these like 152 00:08:19,480 --> 00:08:22,880 Speaker 2: really specific pieces of data and you're asking a computer 153 00:08:22,920 --> 00:08:26,400 Speaker 2: to identify them. What transfer learning is is instead of 154 00:08:26,720 --> 00:08:30,400 Speaker 2: having really specific data, you just give it all the data, right, 155 00:08:30,440 --> 00:08:32,680 Speaker 2: You just give it like the entire Internet. That's what 156 00:08:32,720 --> 00:08:35,360 Speaker 2: open ai I did, and you try to get it to, say, 157 00:08:35,600 --> 00:08:38,920 Speaker 2: finish a paragraph. And the idea is, if you can 158 00:08:38,960 --> 00:08:41,240 Speaker 2: do that, it might be able to do other things 159 00:08:41,720 --> 00:08:44,640 Speaker 2: because in addition to English on the Internet, there's also 160 00:08:44,679 --> 00:08:46,800 Speaker 2: a lot of computer code. So the first thing that 161 00:08:46,840 --> 00:08:49,960 Speaker 2: happened with open ai and with the Microsoft partnership is 162 00:08:50,000 --> 00:08:53,480 Speaker 2: they figured out that actually, this technology that can complete 163 00:08:53,520 --> 00:08:57,320 Speaker 2: sentences for people can also complete computer code. And that's 164 00:08:57,320 --> 00:08:59,360 Speaker 2: not because they asked it to be able to do that. 165 00:08:59,400 --> 00:09:01,600 Speaker 2: It's just because there's a lot of computer code on 166 00:09:01,640 --> 00:09:04,800 Speaker 2: the Internet, and the same basic process that can be 167 00:09:04,920 --> 00:09:07,679 Speaker 2: used to generate English can be used to generate you know, 168 00:09:07,800 --> 00:09:10,480 Speaker 2: C plus plus. And so that's the idea that you 169 00:09:10,520 --> 00:09:14,160 Speaker 2: could start with something like just write sentences and then 170 00:09:14,200 --> 00:09:17,160 Speaker 2: you could say, I don't know, design a road plan 171 00:09:17,480 --> 00:09:20,320 Speaker 2: or design an app or when you talk to the 172 00:09:20,480 --> 00:09:23,120 Speaker 2: like apocalyptic people would be like design a plan to 173 00:09:23,120 --> 00:09:24,880 Speaker 2: take over the world or something like that. And that's 174 00:09:24,920 --> 00:09:27,040 Speaker 2: what the people who are really worried about open ai 175 00:09:27,280 --> 00:09:28,920 Speaker 2: and similar technologies are scared of. 176 00:09:30,120 --> 00:09:32,400 Speaker 1: And so the notion is instead of trying to teach 177 00:09:32,440 --> 00:09:35,360 Speaker 1: these models to do something specific, you kind of teach 178 00:09:35,400 --> 00:09:38,280 Speaker 1: them to do everything, and in teaching them to do everything, 179 00:09:38,320 --> 00:09:40,240 Speaker 1: they can also do the specific thing. 180 00:09:40,679 --> 00:09:44,080 Speaker 2: Yeah. Absolutely, And so an example of this so the 181 00:09:44,120 --> 00:09:46,959 Speaker 2: Microsoft deal, Right, they started getting involved with open ai 182 00:09:47,040 --> 00:09:51,239 Speaker 2: in twenty nineteen, and this partnership broadens. Microsoft starts incorporating 183 00:09:51,280 --> 00:09:54,040 Speaker 2: the technology to its products. It invests more money, as 184 00:09:54,040 --> 00:09:56,439 Speaker 2: I said, billions of dollars in open ai, and a 185 00:09:56,440 --> 00:09:58,520 Speaker 2: lot of people are really excited about this. But there 186 00:09:58,520 --> 00:10:01,440 Speaker 2: are some skeptics, you know, one of whom is Bill Gates, 187 00:10:01,480 --> 00:10:04,320 Speaker 2: you know, the founder of Microsoft, who is very interested 188 00:10:04,320 --> 00:10:08,000 Speaker 2: in AI, and for a really long time, Gates was 189 00:10:08,000 --> 00:10:11,440 Speaker 2: skeptical because he saw that it made mistakes, which anyone 190 00:10:11,440 --> 00:10:14,079 Speaker 2: who's used open ai or chat, GPT or any of 191 00:10:14,120 --> 00:10:17,079 Speaker 2: these services knows. Right, Gates would be talking to the bot, 192 00:10:17,160 --> 00:10:19,880 Speaker 2: right and it would say somebody's in Cleveland in one sentence, 193 00:10:20,080 --> 00:10:22,000 Speaker 2: and the next sentence, they'd be in a different city, 194 00:10:22,040 --> 00:10:24,480 Speaker 2: you know, Seattle, and it was like the bot didn't 195 00:10:24,679 --> 00:10:27,800 Speaker 2: know what was actually going on. And what he said 196 00:10:27,800 --> 00:10:31,280 Speaker 2: to Sam Altman was that I will believe this is 197 00:10:31,400 --> 00:10:35,040 Speaker 2: good if it can pass the ap bio exam. He 198 00:10:35,080 --> 00:10:39,040 Speaker 2: saw that as a test of does this thing actually understand? 199 00:10:39,400 --> 00:10:42,520 Speaker 2: And I talked to Altman about this because I was wondering, like, so, 200 00:10:42,640 --> 00:10:45,640 Speaker 2: did you like feeded a bunch of bio exams to 201 00:10:45,640 --> 00:10:48,400 Speaker 2: try to teach it? And he was like, no, it's 202 00:10:48,480 --> 00:10:50,480 Speaker 2: just that there's a lot of biology on the internet. 203 00:10:50,800 --> 00:10:53,240 Speaker 2: And they had this demo at Bill Gates' house. Bill 204 00:10:53,240 --> 00:10:56,120 Speaker 2: Gates lives in this you know, palatial you know house 205 00:10:56,200 --> 00:11:00,560 Speaker 2: on Lake Washington in the Seattle area, and the Open 206 00:11:00,600 --> 00:11:03,959 Speaker 2: Ai people showed off this early demo of what was 207 00:11:04,000 --> 00:11:07,360 Speaker 2: called GPT four, the latest model, and it was able 208 00:11:07,400 --> 00:11:10,920 Speaker 2: to pass the bio exam, including the essay portion, and 209 00:11:11,000 --> 00:11:12,560 Speaker 2: so Gates was kind of impressed by that. But the 210 00:11:12,600 --> 00:11:14,920 Speaker 2: thing that really kind of brought Gates and the other 211 00:11:14,960 --> 00:11:18,920 Speaker 2: Microsoft executives around, it's the same thing that impressed everybody else. 212 00:11:19,120 --> 00:11:22,360 Speaker 2: It's that after the demo, the open Ai CEO Greg 213 00:11:22,440 --> 00:11:26,040 Speaker 2: Brockman got up on a keyboard and just started asking questions, 214 00:11:26,080 --> 00:11:28,280 Speaker 2: said what do you want to know? And Gates and 215 00:11:28,320 --> 00:11:30,480 Speaker 2: the other executives just started throwing things out, trying to 216 00:11:30,520 --> 00:11:33,640 Speaker 2: stump it. Gates, you know, is really interested in public health, 217 00:11:33,920 --> 00:11:36,920 Speaker 2: interested in, you know, issues of childhood illness, says, how 218 00:11:36,960 --> 00:11:39,160 Speaker 2: would you talk to the parents of a sick child? 219 00:11:39,240 --> 00:11:41,400 Speaker 2: The open ai bought, you know, spits out a response 220 00:11:41,400 --> 00:11:43,640 Speaker 2: and Gates looks at it and says, you know, this 221 00:11:43,679 --> 00:11:46,640 Speaker 2: is more empathetic than I would be. And for him, 222 00:11:46,640 --> 00:11:48,960 Speaker 2: that was the moment, right, that was the moment where 223 00:11:49,000 --> 00:11:54,080 Speaker 2: he thought, Okay, actually this is ready to be a product. 224 00:11:54,200 --> 00:11:56,240 Speaker 2: This isn't just some kind of research experiment. 225 00:11:57,480 --> 00:12:01,680 Speaker 1: So Bill Gates is now on board. What happened next, 226 00:12:02,200 --> 00:12:14,559 Speaker 1: that's after the break So Max, they persuaded Bill Gates, 227 00:12:14,760 --> 00:12:17,200 Speaker 1: who even though he's not on the board anymore, is 228 00:12:17,240 --> 00:12:20,160 Speaker 1: still a key advisor as the founder of the company. 229 00:12:20,600 --> 00:12:23,560 Speaker 2: What happens then, I mean, what's crazy here is how 230 00:12:23,640 --> 00:12:26,680 Speaker 2: quickly this all happens. So that meeting with Gates was 231 00:12:26,720 --> 00:12:29,600 Speaker 2: in summer of twenty twenty two, so less than a 232 00:12:29,640 --> 00:12:33,800 Speaker 2: year ago. As OpenAI and Microsoft start negotiating and as 233 00:12:34,040 --> 00:12:36,800 Speaker 2: the partnership deepens, A lot of this falls on this 234 00:12:36,920 --> 00:12:40,480 Speaker 2: Microsoft executive named Kevin Scott, who Dina Bass, my co writer, 235 00:12:40,600 --> 00:12:42,600 Speaker 2: and I spent a lot of time with for this story. 236 00:12:43,080 --> 00:12:46,480 Speaker 2: Kevin Scott, He's the chief technology officer. He's a really 237 00:12:46,480 --> 00:12:49,600 Speaker 2: interesting guy and sort of an odd duck in this 238 00:12:49,760 --> 00:12:54,160 Speaker 2: world of basically Stanford and Carnegie Mellon and cal Tech PhDs. 239 00:12:54,600 --> 00:12:58,800 Speaker 2: He's from rural Virginia, essentially from Appalachia, went to the 240 00:12:58,880 --> 00:13:03,839 Speaker 2: University of Lynchburg, basically small local Christian college. Eventually kind 241 00:13:03,840 --> 00:13:07,280 Speaker 2: of works his way, you know, into a UVA PhD program, 242 00:13:07,320 --> 00:13:10,440 Speaker 2: winds up at Google, is a huge success at Google, 243 00:13:10,840 --> 00:13:12,920 Speaker 2: and you know, eventually finds the way to become the 244 00:13:13,000 --> 00:13:17,200 Speaker 2: chief Technology officer of Microsoft, where Nadella basically makes it 245 00:13:17,240 --> 00:13:20,800 Speaker 2: his mission to kind of get all these AI programs 246 00:13:20,840 --> 00:13:23,040 Speaker 2: that are kind of all over the place and put 247 00:13:23,080 --> 00:13:25,400 Speaker 2: them in some kind of order. So the funny thing 248 00:13:25,400 --> 00:13:31,319 Speaker 2: about Microsoft has done some really great AI research. They 249 00:13:31,360 --> 00:13:34,600 Speaker 2: also have had some sort of infamous AI products. I 250 00:13:34,600 --> 00:13:39,040 Speaker 2: think probably the first chatbot that anyone ever used, or 251 00:13:39,080 --> 00:13:42,960 Speaker 2: most people used, was Clippy. Most people encountered Clippy, you 252 00:13:42,960 --> 00:13:45,600 Speaker 2: know in Microsoft Word where you'd start typing something, you 253 00:13:45,640 --> 00:13:47,760 Speaker 2: type the word deer, and then all of a sudden, 254 00:13:47,760 --> 00:13:51,440 Speaker 2: this little, you know, anthropomorphic paper clip with gigantic googly 255 00:13:51,520 --> 00:13:55,760 Speaker 2: eyes would come there and sort of like mansplain letters 256 00:13:55,760 --> 00:13:57,839 Speaker 2: to you. It'd be like, well, it looks like you're 257 00:13:57,840 --> 00:13:59,839 Speaker 2: writing deer, but do you need help writing a letter? 258 00:13:59,880 --> 00:14:03,960 Speaker 2: And he would sort of essentially interrupt your work, and 259 00:14:04,160 --> 00:14:06,680 Speaker 2: it became this kind of like gag. There are all 260 00:14:06,720 --> 00:14:09,560 Speaker 2: these funny memes on the Internet of of you know, Clippy, 261 00:14:09,760 --> 00:14:12,520 Speaker 2: you know, interrupting things and just kind of like the 262 00:14:12,559 --> 00:14:16,640 Speaker 2: weird weakness of what otherwise is like an amazing piece 263 00:14:16,679 --> 00:14:18,840 Speaker 2: of software, you know, Microsoft Word. People kind of make 264 00:14:18,880 --> 00:14:20,720 Speaker 2: fun of it sometimes, but like, right, it's one of 265 00:14:20,760 --> 00:14:23,040 Speaker 2: the most successful piece of software of all time, and 266 00:14:23,120 --> 00:14:25,360 Speaker 2: they just stuck a talking paper clip in it for 267 00:14:25,400 --> 00:14:30,360 Speaker 2: no particularly good reason. The other one Microsoft chatbot AI 268 00:14:30,400 --> 00:14:33,640 Speaker 2: thing was this thing called Tay, which happened in twenty sixteen. 269 00:14:34,000 --> 00:14:36,520 Speaker 2: Tay was sort of an update to Clippy where they 270 00:14:36,600 --> 00:14:39,960 Speaker 2: wanted to have a bot that would sort of learn 271 00:14:40,040 --> 00:14:43,960 Speaker 2: from Twitter and would sound like a typical teenager, and 272 00:14:44,080 --> 00:14:48,120 Speaker 2: within about i don't know, twenty four hours, Tay had 273 00:14:48,160 --> 00:14:51,160 Speaker 2: been deluged with kind of what you might imagine is 274 00:14:51,200 --> 00:14:54,560 Speaker 2: on Twitter. She went from being like a nice, helpful 275 00:14:54,600 --> 00:14:59,080 Speaker 2: teenager to just like sounding like the most obnoxious person 276 00:14:59,160 --> 00:15:01,360 Speaker 2: on you know, on four Chan or on Reddit or whatever, 277 00:15:01,640 --> 00:15:02,520 Speaker 2: and they had to pull it. 278 00:15:04,640 --> 00:15:07,400 Speaker 1: What does Microsoft say about Clippy and Tay when they 279 00:15:07,480 --> 00:15:11,520 Speaker 1: look back on their early experiments with this technology. 280 00:15:12,720 --> 00:15:15,960 Speaker 2: From Microsoft's point of view, you know, Tay was done 281 00:15:15,960 --> 00:15:19,680 Speaker 2: in by people who were manipulating Tay, sending it hate speech, 282 00:15:19,760 --> 00:15:21,440 Speaker 2: you know, trying to get it to say the worst 283 00:15:21,440 --> 00:15:25,200 Speaker 2: possible things. And so Kevin Scott has kind of the 284 00:15:25,240 --> 00:15:28,560 Speaker 2: weight of all this kind of history of Microsoft's past 285 00:15:28,920 --> 00:15:32,960 Speaker 2: bad efforts, as well as this research program that was 286 00:15:33,080 --> 00:15:35,320 Speaker 2: good but just kind of unfocused. 287 00:15:36,400 --> 00:15:40,640 Speaker 1: So now Kevin Scott has this mandate to take this 288 00:15:40,840 --> 00:15:45,800 Speaker 1: very powerful AI chat generator and make use of it, 289 00:15:45,840 --> 00:15:47,480 Speaker 1: and he's got a lot of money to do it. 290 00:15:47,840 --> 00:15:48,840 Speaker 1: And what did they. 291 00:15:48,720 --> 00:15:52,000 Speaker 2: Do before this the meeting with Bill Gates? They had 292 00:15:52,040 --> 00:15:55,480 Speaker 2: had one product, which was GitHub. So GitHub is a 293 00:15:55,480 --> 00:15:58,640 Speaker 2: product for software coders. And they'd sort of figured out 294 00:15:59,040 --> 00:16:02,800 Speaker 2: that at GPT, or really GPT three, which is the 295 00:16:02,880 --> 00:16:06,200 Speaker 2: underlying technology, that in addition to being able to finish sentences, 296 00:16:06,400 --> 00:16:08,800 Speaker 2: it could finish code. So they created this thing called 297 00:16:08,880 --> 00:16:12,240 Speaker 2: Copilot for GitHub and it's like auto complete, you know, 298 00:16:12,280 --> 00:16:15,240 Speaker 2: autocomplete on your text message where you're typing a message 299 00:16:15,280 --> 00:16:17,520 Speaker 2: and it sort of tries to guess what you're about 300 00:16:17,560 --> 00:16:20,240 Speaker 2: to say. It was like that except for software code, 301 00:16:20,280 --> 00:16:22,560 Speaker 2: and it was pretty successful. They were charging money for 302 00:16:22,600 --> 00:16:24,840 Speaker 2: it. It costs, you know, between ten and twenty dollars a 303 00:16:24,880 --> 00:16:28,040 Speaker 2: month per person, so real money. And that was it. 304 00:16:28,040 --> 00:16:30,440 Speaker 2: That's all they had. And for the next one, as 305 00:16:30,480 --> 00:16:33,760 Speaker 2: part of this big push, they chose Bing. It's kind 306 00:16:33,800 --> 00:16:36,040 Speaker 2: of like the Clippy of search engines, right, it's this 307 00:16:36,240 --> 00:16:40,800 Speaker 2: like kind of goofy, unsuccessful bit of a head scratcher 308 00:16:41,040 --> 00:16:43,440 Speaker 2: of a competitor to Google. You know a lot of 309 00:16:43,440 --> 00:16:45,800 Speaker 2: people like Bing, but it's never really taken off. I 310 00:16:45,800 --> 00:16:48,880 Speaker 2: think partly because Google just dominates the market, and I 311 00:16:48,880 --> 00:16:52,000 Speaker 2: think they just thought, well, search is a really big business. 312 00:16:52,360 --> 00:16:56,360 Speaker 2: These chat bots are potentially pretty useful for search because 313 00:16:56,360 --> 00:16:58,720 Speaker 2: you could ask it instead of searching for a restaurant, 314 00:16:58,760 --> 00:17:01,760 Speaker 2: you'd say, what's the best restaurant? And being the stakes 315 00:17:01,800 --> 00:17:04,360 Speaker 2: are low, right, if it's terrible, they're not that many 316 00:17:04,400 --> 00:17:06,440 Speaker 2: BING users out there to be up in arms. 317 00:17:07,119 --> 00:17:11,920 Speaker 1: So what happened when they put this technology into being 318 00:17:12,640 --> 00:17:13,480 Speaker 1: in February. 319 00:17:13,880 --> 00:17:16,439 Speaker 2: You know, Sati Ndella has an event. He gets up 320 00:17:16,480 --> 00:17:19,000 Speaker 2: on stage and he shows off this technology and everyone 321 00:17:19,080 --> 00:17:22,399 Speaker 2: is really impressed. Right. Bing is based on the most 322 00:17:22,440 --> 00:17:26,600 Speaker 2: advanced version of open aiyes model, it's called GPT four. 323 00:17:27,040 --> 00:17:29,760 Speaker 2: During this demo they showed it you know, writing letters 324 00:17:29,800 --> 00:17:31,560 Speaker 2: and planning menus and things like that. 325 00:17:32,080 --> 00:17:35,399 Speaker 1: So you're not just using being to say, you know, 326 00:17:35,760 --> 00:17:40,919 Speaker 1: find me this restaurant's website. You're actually asking you questions 327 00:17:40,960 --> 00:17:44,240 Speaker 1: and then it retrieves it. So it's searching for answers, 328 00:17:44,280 --> 00:17:46,240 Speaker 1: not just for results. 329 00:17:47,040 --> 00:17:49,680 Speaker 2: Yes, and this is really exciting. I mean, first of all, 330 00:17:49,680 --> 00:17:53,280 Speaker 2: it's exciting because it could be potentially easier, right, instead 331 00:17:53,320 --> 00:17:56,520 Speaker 2: of spending you know, an hour on Google searching for 332 00:17:57,040 --> 00:18:00,240 Speaker 2: restaurants in Paris and hotels in Paris and activity in 333 00:18:00,280 --> 00:18:02,880 Speaker 2: Paris to make my Parris itinerary, you could just ask 334 00:18:02,960 --> 00:18:05,080 Speaker 2: a bot to make you one, right, and they'll do it. 335 00:18:05,240 --> 00:18:07,760 Speaker 2: They'll give you something, you know, within I don't know, 336 00:18:07,840 --> 00:18:11,480 Speaker 2: ninety seconds or whatever, And that's really exciting. Potentially makes 337 00:18:11,520 --> 00:18:14,159 Speaker 2: things easier, I guess if you're traveling. But from a 338 00:18:14,200 --> 00:18:19,280 Speaker 2: business perspective, that's like a huge moment because it could 339 00:18:19,320 --> 00:18:22,400 Speaker 2: potentially you know, break a lot of the search business 340 00:18:22,520 --> 00:18:24,879 Speaker 2: because if you don't spend that hour on Google, if 341 00:18:24,920 --> 00:18:27,280 Speaker 2: you're not clicking all these links, Google is not able 342 00:18:27,280 --> 00:18:29,479 Speaker 2: to put ads in front of you. So you know, 343 00:18:29,600 --> 00:18:32,920 Speaker 2: immediately you saw Google stock fall, you saw Microsoft stock 344 00:18:32,960 --> 00:18:36,840 Speaker 2: go up, and then people started playing around with Being. 345 00:18:37,480 --> 00:18:41,040 Speaker 2: You had all these reporters and Twitter users spending time 346 00:18:41,080 --> 00:18:44,280 Speaker 2: with Being and getting it to do some weird things. 347 00:18:44,800 --> 00:18:47,080 Speaker 2: The most famous one is Kevin Russ, who's a tech 348 00:18:47,080 --> 00:18:49,479 Speaker 2: columnist of the New York Times, had this like two 349 00:18:49,560 --> 00:18:53,200 Speaker 2: hour conversation with Being about love where Being became an 350 00:18:53,200 --> 00:18:55,879 Speaker 2: alter ego called Sydney. I mean, it sounds insane when 351 00:18:56,000 --> 00:18:57,679 Speaker 2: when I say it and told him to leave his 352 00:18:57,720 --> 00:19:00,280 Speaker 2: wife and said no, no, you love me and got 353 00:19:00,320 --> 00:19:03,600 Speaker 2: really aggressive and just kind of flipped out. You had 354 00:19:03,640 --> 00:19:07,440 Speaker 2: another sort of famous tech analyst, Ben Thompson, got bing 355 00:19:07,600 --> 00:19:10,320 Speaker 2: to adopt a different alter ego, venom As, kind of 356 00:19:10,359 --> 00:19:13,480 Speaker 2: like an evil version of Sydney, kind of fantasizing about 357 00:19:13,520 --> 00:19:17,399 Speaker 2: how you would extract revenge on his enemies. Being told 358 00:19:17,440 --> 00:19:19,639 Speaker 2: the Verge, you know, it's a tech publication that it 359 00:19:19,720 --> 00:19:23,199 Speaker 2: was spying on Microsoft employees using their webcams. Like, so 360 00:19:23,280 --> 00:19:25,679 Speaker 2: not only making something up, but making something like you know, 361 00:19:25,720 --> 00:19:28,960 Speaker 2: potentially pretty damaging to Microsoft and so on. 362 00:19:29,359 --> 00:19:34,119 Speaker 1: And Max, what is Microsoft say about these encounters you 363 00:19:34,240 --> 00:19:37,160 Speaker 1: described that people had with being so. 364 00:19:37,119 --> 00:19:41,000 Speaker 2: From Microsoft's perspective, the problem wasn't the software, it was 365 00:19:41,160 --> 00:19:43,920 Speaker 2: the people who were using it. They were, as CTO 366 00:19:44,119 --> 00:19:46,720 Speaker 2: Kevin Scott said to us, basically trying to goat it 367 00:19:46,760 --> 00:19:50,360 Speaker 2: in to say the worst possible things. They just they're 368 00:19:50,400 --> 00:19:52,960 Speaker 2: moving ahead. They're not actually stopping. I mean, they made 369 00:19:52,960 --> 00:19:55,800 Speaker 2: some tweaks to how BING works, but it's not like 370 00:19:55,840 --> 00:19:57,880 Speaker 2: they're not going to bring this stuff to office. They're 371 00:19:57,920 --> 00:19:59,359 Speaker 2: still going to bring it to office. They're going to 372 00:19:59,400 --> 00:20:02,440 Speaker 2: bring it to who what is like probably the biggest 373 00:20:02,480 --> 00:20:05,240 Speaker 2: and most important piece of software in the world, which 374 00:20:05,280 --> 00:20:08,280 Speaker 2: is you know, these productivity apps that Microsoft sells in 375 00:20:08,320 --> 00:20:14,399 Speaker 2: other words, Microsoft Word, Microsoft Excel, Microsoft PowerPoint, And this 376 00:20:14,520 --> 00:20:17,360 Speaker 2: is already happening like inside of certain companies and stuff. 377 00:20:17,480 --> 00:20:19,600 Speaker 2: You will be able to be, you know, working on 378 00:20:19,640 --> 00:20:22,000 Speaker 2: a chart and Excel and you'll be able to ask 379 00:20:22,080 --> 00:20:25,600 Speaker 2: the copilot to make a visualization of that, or you'll 380 00:20:25,600 --> 00:20:27,560 Speaker 2: be able to, you know, as I said, like have 381 00:20:27,640 --> 00:20:30,840 Speaker 2: it essentially create rough versions of your PowerPoint slides. For 382 00:20:30,880 --> 00:20:34,280 Speaker 2: a presentation, and the idea here is not to have 383 00:20:34,320 --> 00:20:37,280 Speaker 2: it sort of replace your work. It's not going to 384 00:20:37,359 --> 00:20:39,840 Speaker 2: do your presentation for you, but it just might make 385 00:20:39,880 --> 00:20:43,080 Speaker 2: it a lot easier. The way that a Microsoft executive 386 00:20:43,280 --> 00:20:46,040 Speaker 2: who works on these products told us is it's like 387 00:20:46,359 --> 00:20:49,000 Speaker 2: it might do maybe half of the work right. You're 388 00:20:49,000 --> 00:20:50,439 Speaker 2: still going to have to keep an eye on it. 389 00:20:50,480 --> 00:20:52,240 Speaker 2: You're going to have to make sure it doesn't make 390 00:20:52,280 --> 00:20:54,720 Speaker 2: any huge mistakes, it doesn't make anything up, and it's 391 00:20:54,720 --> 00:20:56,480 Speaker 2: not going to do everything, but it hopefully will make 392 00:20:56,520 --> 00:20:58,040 Speaker 2: things like easier. 393 00:20:58,359 --> 00:21:02,639 Speaker 1: When we come back. Microsoft has big future plans for AI, 394 00:21:11,920 --> 00:21:14,440 Speaker 1: so now max. In a really short amount of time, 395 00:21:14,560 --> 00:21:20,080 Speaker 1: this went from novelty to a core part of Microsoft's 396 00:21:20,119 --> 00:21:23,280 Speaker 1: business and has really leapfroged the company in a lot 397 00:21:23,280 --> 00:21:25,840 Speaker 1: of ways ahead of some of its biggest competitors. 398 00:21:26,280 --> 00:21:28,040 Speaker 2: If you went back a couple of years and you 399 00:21:28,080 --> 00:21:30,480 Speaker 2: asked people at Silicon Valley, like, who's in the lead 400 00:21:30,600 --> 00:21:33,720 Speaker 2: in AI, they would have said Google. And I think 401 00:21:33,840 --> 00:21:37,399 Speaker 2: everyone agrees now that Microsoft is either kind of like 402 00:21:37,560 --> 00:21:40,400 Speaker 2: in the lead or has at least gotten to parody 403 00:21:40,440 --> 00:21:44,320 Speaker 2: with Google, which is huge achievement because Google Larry Page, 404 00:21:44,320 --> 00:21:46,840 Speaker 2: Sergey Brin. I mean, they have been really focused on 405 00:21:46,880 --> 00:21:48,520 Speaker 2: this for a very long time. We know about the 406 00:21:48,520 --> 00:21:51,080 Speaker 2: Google self driving car. There's this thing called deep Mind, 407 00:21:51,359 --> 00:21:54,080 Speaker 2: which before open Ai was the hot you know, research lab, 408 00:21:54,080 --> 00:21:56,639 Speaker 2: which is owned by Google. And on the business front, 409 00:21:56,680 --> 00:21:58,119 Speaker 2: I mean, they are looking like they are in a 410 00:21:58,200 --> 00:22:01,320 Speaker 2: very good position. This isn't to say that this couldn't 411 00:22:01,359 --> 00:22:03,840 Speaker 2: like all come crashing down or I think it's very 412 00:22:03,880 --> 00:22:07,560 Speaker 2: possible that these technologies just don't work as well as 413 00:22:07,600 --> 00:22:10,480 Speaker 2: they're hoping and aren't as useful as people think. But 414 00:22:10,720 --> 00:22:14,760 Speaker 2: here's the thing. Microsoft is very very good at extracting 415 00:22:14,840 --> 00:22:20,000 Speaker 2: money out of software. Right for basically two decades, there 416 00:22:20,040 --> 00:22:24,439 Speaker 2: have been free alternatives to Microsoft Office, Word Excel. You 417 00:22:24,480 --> 00:22:26,280 Speaker 2: can get a free version of them. And you've been 418 00:22:26,280 --> 00:22:29,160 Speaker 2: able to get a free version of these things for decades, 419 00:22:29,400 --> 00:22:32,240 Speaker 2: and yet they're pulling in, you know, something like forty 420 00:22:32,240 --> 00:22:35,600 Speaker 2: five billion dollars a year on licensing fees on these technologies. 421 00:22:35,920 --> 00:22:39,720 Speaker 2: And so their plan here with these chatbots is essentially 422 00:22:39,840 --> 00:22:43,520 Speaker 2: use them to drive more revenue to that thing. So 423 00:22:43,760 --> 00:22:45,680 Speaker 2: from now on, if you're you're gonna pay your ninety 424 00:22:45,720 --> 00:22:49,159 Speaker 2: nine dollars a year for your office subscription, if they 425 00:22:49,160 --> 00:22:51,480 Speaker 2: price it like GitHub, it's going to be you know, 426 00:22:51,520 --> 00:22:53,960 Speaker 2: a ten dollars a month. Add on. A lot of 427 00:22:54,000 --> 00:22:56,480 Speaker 2: people and especially a lot of businesses are going to 428 00:22:56,560 --> 00:22:58,720 Speaker 2: feel like they need this and they're going to pay 429 00:22:58,720 --> 00:23:01,400 Speaker 2: for it. And Microsoft, as I say, is very good 430 00:23:01,440 --> 00:23:03,720 Speaker 2: at convincing people to pay for its software. It's kind 431 00:23:03,720 --> 00:23:08,439 Speaker 2: of like what the company is almost best at. The 432 00:23:08,520 --> 00:23:11,919 Speaker 2: other thing that they have is this supercomputer that I 433 00:23:11,960 --> 00:23:15,040 Speaker 2: brought up, where it's not just that Microsoft can put 434 00:23:15,080 --> 00:23:18,200 Speaker 2: open AI in all its apps, it's that the entire 435 00:23:18,240 --> 00:23:22,480 Speaker 2: world of business is attempting to integrate AI into their products, 436 00:23:22,600 --> 00:23:25,359 Speaker 2: not just Microsoft products, but other products, and they're going 437 00:23:25,440 --> 00:23:27,520 Speaker 2: to need computers to do that. And Microsoft right now 438 00:23:27,560 --> 00:23:31,080 Speaker 2: is in the best position and position to sell those services, 439 00:23:31,280 --> 00:23:34,800 Speaker 2: the kind of cloud services to everyone. And so when 440 00:23:34,840 --> 00:23:37,159 Speaker 2: you add those two things up, you're talking about a 441 00:23:37,240 --> 00:23:40,040 Speaker 2: huge amount of revenue. We mentioned this in the story, 442 00:23:40,080 --> 00:23:42,840 Speaker 2: but one analyst has a model, which I think is debatable, 443 00:23:42,880 --> 00:23:44,919 Speaker 2: but this is like a legitimate model where he's talking 444 00:23:44,920 --> 00:23:49,760 Speaker 2: about by twenty twenty seven, somewhere between ten billion at 445 00:23:49,760 --> 00:23:52,440 Speaker 2: the low end and one hundred billion at the high 446 00:23:52,560 --> 00:23:55,119 Speaker 2: end of additional revenue to this company. So that's like 447 00:23:55,359 --> 00:23:57,879 Speaker 2: if they get to one hundred billion, that would be 448 00:23:57,920 --> 00:24:00,760 Speaker 2: like it's like adding three netflixes to a company that's 449 00:24:00,800 --> 00:24:03,560 Speaker 2: already the second largest in the world by market cap. 450 00:24:04,000 --> 00:24:06,680 Speaker 2: I do think it's important to say that the other 451 00:24:06,720 --> 00:24:10,080 Speaker 2: companies aren't standing still, right. They see the same opportunities 452 00:24:10,080 --> 00:24:13,359 Speaker 2: at Microsoft. See so Google is frantically trying to put 453 00:24:13,440 --> 00:24:16,160 Speaker 2: this technology into their apps, and I think there's reason 454 00:24:16,200 --> 00:24:18,240 Speaker 2: to think Microsoft is in a better position because of 455 00:24:18,240 --> 00:24:21,120 Speaker 2: the commercial factors I brought up. But you could imagine 456 00:24:21,119 --> 00:24:24,160 Speaker 2: a situation where in a year from now or two 457 00:24:24,200 --> 00:24:26,480 Speaker 2: years from now, people are a little less excited about 458 00:24:26,520 --> 00:24:29,199 Speaker 2: it all of a sudden, you know, you maybe you 459 00:24:29,200 --> 00:24:32,000 Speaker 2: don't want to spend extra money for a chat GBT thing, 460 00:24:32,400 --> 00:24:34,919 Speaker 2: where then it becomes a little more challenging for Microsoft. 461 00:24:35,160 --> 00:24:37,840 Speaker 2: But even then, I think it's still going to be 462 00:24:37,880 --> 00:24:41,120 Speaker 2: a substantial amount of revenue for this company, just because 463 00:24:41,160 --> 00:24:43,600 Speaker 2: they have such a big footprint and they're so far 464 00:24:43,640 --> 00:24:46,920 Speaker 2: ahead of everyone else. Max. 465 00:24:46,920 --> 00:24:51,360 Speaker 1: What does Microsoft say about these growing concerns about sort 466 00:24:51,359 --> 00:24:54,400 Speaker 1: of doomsday scenarios having to do with AI. 467 00:24:54,840 --> 00:24:57,000 Speaker 2: They do not see it as a thing that's going 468 00:24:57,000 --> 00:25:00,320 Speaker 2: to end the world. They see a computing platform. The 469 00:25:00,359 --> 00:25:03,280 Speaker 2: Microsoft people right, are not super focused on the end 470 00:25:03,280 --> 00:25:05,480 Speaker 2: the world questions, which sort of makes sense when you're 471 00:25:05,480 --> 00:25:08,359 Speaker 2: talking about like a copilot for Excel. It's hard to 472 00:25:08,359 --> 00:25:11,720 Speaker 2: imagine like spell check taking over the world, And so 473 00:25:11,800 --> 00:25:14,359 Speaker 2: what they're really interested in is sort of like productivity. 474 00:25:14,640 --> 00:25:17,720 Speaker 2: How can this make workers better? And what they say 475 00:25:17,760 --> 00:25:20,879 Speaker 2: about these kind of other concerns, they say, like it's copilot, 476 00:25:20,920 --> 00:25:24,479 Speaker 2: it's not autopilot, and it really is not that useful 477 00:25:24,520 --> 00:25:27,400 Speaker 2: as a replacement. It's only useful as a thing that 478 00:25:27,400 --> 00:25:29,600 Speaker 2: can like sort of help you along. I think there 479 00:25:29,600 --> 00:25:33,880 Speaker 2: are questions about how applicable that is, whether it actually 480 00:25:33,960 --> 00:25:37,439 Speaker 2: ends up happening, whether these technologies get used responsibly, But 481 00:25:37,480 --> 00:25:39,400 Speaker 2: that's kind of their point of view that like, as 482 00:25:39,400 --> 00:25:41,840 Speaker 2: long as humans are in the mix, a lot of 483 00:25:41,880 --> 00:25:43,200 Speaker 2: these risks are not that great. 484 00:25:44,240 --> 00:25:46,440 Speaker 1: Max, thanks so much for coming on the show. 485 00:25:47,000 --> 00:25:48,880 Speaker 2: Thanks for keeping this human in the mix. 486 00:25:51,600 --> 00:25:53,440 Speaker 1: Thanks for listening to us here at the Big Take. 487 00:25:53,560 --> 00:25:57,000 Speaker 1: It's a daily podcast from Bloomberg and iHeartRadio. For more 488 00:25:57,040 --> 00:26:01,400 Speaker 1: shows from iHeartRadio, visit the iHeartRadio app podcasts, or wherever 489 00:26:01,440 --> 00:26:04,080 Speaker 1: you listen, and we'd love to hear from you. Email 490 00:26:04,160 --> 00:26:07,720 Speaker 1: us questions or comments to Big Take at Bloomberg dot Net. 491 00:26:08,320 --> 00:26:11,480 Speaker 1: The supervising producer of the Big Take is Vicky Vergalina. 492 00:26:11,880 --> 00:26:16,200 Speaker 1: Our senior producer is Katherine Fink. Rebecca Shasson is our producer. 493 00:26:16,520 --> 00:26:20,480 Speaker 1: Our associate producer is Sam Gebauer. Raphae alum See is 494 00:26:20,520 --> 00:26:24,160 Speaker 1: our engineer. Our original music was composed by Leo Sidrin. 495 00:26:24,520 --> 00:26:27,920 Speaker 1: I'm west Kasova. We'll be back tomorrow with another Big Take.