WEBVTT - Sam Altman Shock Ouster from OpenAI 0:00:01.400 --> 0:00:05.080 From Mahart where Innovation of Money and Power Collie in 0:00:05.200 --> 0:00:10.080 Silicon Valley, NBN. This is Bloomberg Technology with Caroline Hyde 0:00:10.160 --> 0:00:11.080 and Ed Ludlove. 0:00:24.440 --> 0:00:26.840 I'm Caroline Heinde of Bloomberg's World headquarters in New York. 0:00:27.360 --> 0:00:30.760 I'm Ed Ludlow in San Francisco. This is Bloomberg Technology. 0:00:30.880 --> 0:00:31.320 Coming up. 0:00:31.400 --> 0:00:34.120 We take a deep dive into the shock olster of 0:00:34.200 --> 0:00:37.880 Sam Altman from Open Ai. What happened, what lies ahead 0:00:37.880 --> 0:00:40.199 for the AI startup as most of the company's staff 0:00:40.360 --> 0:00:42.520 threatened to depart and. 0:00:42.479 --> 0:00:46.320 A clear win for Microsoft shares of the tech giant rises. 0:00:46.360 --> 0:00:50.680 Analysts see risks contained for the company's AI ambitions as 0:00:50.720 --> 0:00:54.600 Altman joins Microsoft to lead a new advanced AI research team. 0:00:54.760 --> 0:00:57.000 Plus, we'll speak to developers in the field of AI 0:00:57.120 --> 0:00:59.800 and the venture capitalists backing them this hour to discuss 0:00:59.800 --> 0:01:02.320 what the ecosystem looks like for these startups and in 0:01:02.400 --> 0:01:04.080 the upheaval and Open Ai. 0:01:04.319 --> 0:01:07.520 This we speak, there are hundreds of Open Ai staff 0:01:07.760 --> 0:01:10.160 that are taking action. Let's break down the news with 0:01:10.240 --> 0:01:13.840 Bloomberg's Rachel Metz, who is joining me on set in 0:01:13.840 --> 0:01:16.440 San Francisco, having stayed up with me all weekend to 0:01:16.440 --> 0:01:18.559 break this story. What do we know about what's happening 0:01:18.560 --> 0:01:20.720 inside Open AIHQ right now? 0:01:20.760 --> 0:01:24.720 It sounds like the majority of the company's employees have 0:01:24.840 --> 0:01:28.840 said that they will leave if their demands are not there. 0:01:28.959 --> 0:01:32.760 They are really upset about this ousting of Sam Altman 0:01:33.200 --> 0:01:35.760 and then the subsequent resignation of Greg Brockman. 0:01:36.120 --> 0:01:38.280 So the numbers, I think are that five hundred and 0:01:38.360 --> 0:01:42.360 ninety nine out of seven hundred and seventy staff. In 0:01:42.400 --> 0:01:45.160 other words, the vast majority of staff have signed this letter. 0:01:47.520 --> 0:01:49.560 Who side did this staff on here? I mean, this 0:01:49.640 --> 0:01:53.200 is hard to understand that we're talking about everyone other 0:01:53.280 --> 0:01:54.800 than those that have departed already. 0:01:55.680 --> 0:01:58.400 Yeah, pretty much. I mean this What I thought was 0:01:58.440 --> 0:02:02.040 really interesting when I saw this letter early this morning 0:02:02.560 --> 0:02:05.600 is that it makes it very very clear whose side 0:02:05.760 --> 0:02:08.320 most of the employees are on. Like I think before 0:02:08.400 --> 0:02:10.639 there was a sense that a lot of people were 0:02:11.240 --> 0:02:14.200 on Sam Aultman's side, but it wasn't crystal clear. This 0:02:14.360 --> 0:02:18.040 is saying, hey, we almost all of us are very angry, 0:02:18.160 --> 0:02:20.680 so angry we're willing to leave our jobs at what 0:02:20.840 --> 0:02:25.160 has been the world's leading independent AI company. 0:02:25.520 --> 0:02:29.760 Rachel number twelve took my eye on the list of 0:02:29.800 --> 0:02:34.000 those who have signed this letter. It's Ilia Ilia Siskeva, 0:02:34.080 --> 0:02:36.280 who is on the board, who. 0:02:36.160 --> 0:02:36.960 Made the cools. 0:02:37.280 --> 0:02:39.880 We understand through all of your guys reporting made the 0:02:39.919 --> 0:02:43.160 calls to Sam Auntlan and Gregg himself to basically let 0:02:43.160 --> 0:02:45.000 them know that they're no longer going to be at 0:02:45.040 --> 0:02:45.520 the company. 0:02:47.000 --> 0:02:47.240 Yeah. 0:02:47.240 --> 0:02:51.760 That's really a bit confusing, right, I mean at first 0:02:51.840 --> 0:02:54.440 I was like, hmm, but you know, it's been a 0:02:54.480 --> 0:02:58.560 really long day that's lasted three days, So yeah, I 0:02:58.560 --> 0:03:00.720 mean I think there he we did. I think it 0:03:00.760 --> 0:03:06.040 was this morning that he made a mistake, right and 0:03:06.080 --> 0:03:07.960 he I mean it sounds I don't think he wants 0:03:08.120 --> 0:03:12.240 the company to explode. Perhaps he didn't anticipate what the 0:03:12.280 --> 0:03:14.680 consequences would be of the actions he was taking. 0:03:15.080 --> 0:03:18.320 And no wonder we are getting so much views and 0:03:18.960 --> 0:03:22.760 theories being shared ultimately about corporate governance. Will go into 0:03:22.760 --> 0:03:24.880 that much more later in the show, Rachel, But I'm 0:03:24.919 --> 0:03:29.040 interested in also ultimately who's being drafted in to replace it. 0:03:29.240 --> 0:03:33.240 How on earth does the new CEO of open AI 0:03:33.760 --> 0:03:35.840 try to ensure that we don't have this mass departure. 0:03:37.480 --> 0:03:41.320 That is an excellent question. So the person who was 0:03:41.400 --> 0:03:45.520 named very late yesterday as the new CEO by the board, 0:03:45.720 --> 0:03:49.960 Emmett's sheer. He was a co founder of the video 0:03:49.960 --> 0:03:53.320 game streaming service Twitch, which has also of course been 0:03:53.360 --> 0:03:55.040 used to for all kinds of streaming at this point, 0:03:55.800 --> 0:03:58.600 em it is a little bit is more on the 0:03:58.640 --> 0:04:01.720 conservative side of tech chnology development in the sense that 0:04:02.040 --> 0:04:05.280 he is in favor of slowing things down somewhat, which 0:04:05.320 --> 0:04:10.720 is what Ilia had been also pushing. So sort of, 0:04:10.760 --> 0:04:15.040 you know, aligned with the board members and with Ilia, 0:04:15.160 --> 0:04:17.919 He's gonna probably have to do a lot of convincing 0:04:18.120 --> 0:04:20.159 to the staff if he wants to get them to 0:04:20.200 --> 0:04:22.400 stay with him in charge. It's going to be very 0:04:22.400 --> 0:04:25.120 interesting to see what happens over the next I was 0:04:25.120 --> 0:04:26.960 about to say days, but let's say hours. 0:04:27.200 --> 0:04:30.480 Well, let's recap. It's only been forty eight hours essentially, 0:04:30.520 --> 0:04:33.480 maybe more since news broke that Sam Outman have been fired. 0:04:33.600 --> 0:04:36.240 This is the timeline of what happened. You and I 0:04:36.279 --> 0:04:38.240 and all of our colleagues spent most of the weekend 0:04:38.920 --> 0:04:41.839 with everyone else thinking Sam's coming back. This is happening 0:04:42.160 --> 0:04:44.880 based on the reporting since explained to us what actually 0:04:44.920 --> 0:04:46.960 happened and the chronology of events. 0:04:47.600 --> 0:04:49.960 So it how far back do you want me to go? 0:04:50.040 --> 0:04:50.200 Here. 0:04:50.680 --> 0:04:53.480 Gosh, let's say we start with Sam being fired on Friday. 0:04:53.520 --> 0:04:56.880 Okay, on Friday, we were told that Sam had been 0:04:57.040 --> 0:04:59.400 fired by the board, which I think came as a 0:04:59.440 --> 0:05:02.640 shock to probably pretty much anybody except for the board. 0:05:03.760 --> 0:05:06.719 Once that happened, it was a scramble to figure out, well, 0:05:06.760 --> 0:05:10.359 you know, what happened here was did something really bad happen? 0:05:10.520 --> 0:05:15.679 It sounded very mysterious, and information started coming from Open 0:05:15.720 --> 0:05:19.279 Ai executives themselves in the form of memos they were 0:05:19.279 --> 0:05:22.440 sending internally that the board has not told us of 0:05:22.440 --> 0:05:25.000 any malfeasance, for instance, which I think was the word 0:05:25.080 --> 0:05:29.039 used in a memo that bred Lightcap sent out to Yes, yeah, 0:05:29.200 --> 0:05:33.159 thank you. So then it started to sound like Mira Marati, 0:05:33.200 --> 0:05:35.680 who was named interim CEO. She had been the chief 0:05:35.720 --> 0:05:40.400 technology officer and increasingly a prominent figure at the company 0:05:40.480 --> 0:05:45.880 alongside Sam. It sounded like perhaps she was going to 0:05:46.520 --> 0:05:51.479 rehire Sam Altman and Greg Brockman, who had been also 0:05:51.560 --> 0:05:55.599 at the company and quit after Sam was ousted. Then, 0:05:55.839 --> 0:05:58.240 in a last minute twist, it sounded like Emmett Sheer 0:05:58.320 --> 0:06:01.799 was hired, and now everybody assigned this letter, or not everybody, 0:06:01.800 --> 0:06:05.000 but most people assigned this letter saying, oh, we're going 0:06:05.080 --> 0:06:07.839 to join this company that Microsoft is now creating, that 0:06:07.960 --> 0:06:11.359 Sam and Gregor are going to run. So if you 0:06:11.360 --> 0:06:13.840 you know, if you don't bring them back essentially So, 0:06:14.279 --> 0:06:15.800 I don't know about you, guys, but I keep just 0:06:15.880 --> 0:06:17.440 kind of turning left and right. I'm not really sure 0:06:17.480 --> 0:06:18.599 what's going to happen next year. 0:06:18.760 --> 0:06:22.120 Whiplash and a whole lot of discussion on what is 0:06:22.120 --> 0:06:24.719 meant to be holiday week. Rachel Mets, We thank you 0:06:24.800 --> 0:06:26.920 so much making time for us when I know you've 0:06:26.920 --> 0:06:29.480 just been at the grindstone with your reporting. We thank 0:06:29.520 --> 0:06:32.040 you so much. And look we just finished where well, 0:06:32.160 --> 0:06:34.960 Rachel was discussing where next for Microsoft as it sets 0:06:35.040 --> 0:06:37.680 up this new AI focus point, what is the impact 0:06:37.760 --> 0:06:40.600 on the publicly traded company, what of Satyr and Nadella 0:06:40.880 --> 0:06:44.520 and really his ability to turn this story around in 0:06:44.560 --> 0:06:47.360 the final few hours joining us as Jeffrey's senior Internet analyst, 0:06:47.400 --> 0:06:49.960 brend Phil Brent, You've got a rating on Microsoft four 0:06:50.000 --> 0:06:53.839 hundred dollars price target? Is this a positive as it stands. 0:06:55.640 --> 0:06:58.000 As a standard, It's a win win for Microsoft. I 0:06:58.000 --> 0:07:00.719 mean they get the number one up, up and coming 0:07:00.800 --> 0:07:06.159 individual and tech leading AI journey. And then secondarily, they've 0:07:06.160 --> 0:07:08.440 been talking about this for a long time. How do 0:07:08.520 --> 0:07:11.240 they reduce their independence on one vendor like open ai 0:07:11.360 --> 0:07:14.920 and spread their AI bats across open source, across their 0:07:14.960 --> 0:07:19.280 own languages, across other languages. And we think that Altman 0:07:19.320 --> 0:07:23.080 can probably steer that chip in a really good way. 0:07:24.120 --> 0:07:28.520 From internally at Microsoft, obviously, from a hiring perspective, if 0:07:28.600 --> 0:07:30.720 you've got over five hundred people signing a petition that 0:07:30.760 --> 0:07:33.480 feel this strong, I've never seen this in my coverage 0:07:33.480 --> 0:07:36.280 at Tech in twenty years, or literally five hundred people 0:07:36.360 --> 0:07:37.120 sign a letter. 0:07:37.400 --> 0:07:40.120 Internally, you've never seen this at Tech ever. 0:07:40.920 --> 0:07:44.560 So if they feel that strongly about him, and he says, hey, 0:07:44.600 --> 0:07:47.120 would you like to come over here to Microsoft? Like, 0:07:47.400 --> 0:07:49.640 what's the tailent pool? Microsoft can to look like? You 0:07:49.680 --> 0:07:52.000 know two months from now, it's going to be insane. 0:07:52.040 --> 0:07:56.800 Well, friend. The question that I hear out in the 0:07:56.800 --> 0:08:00.720 market is will Microsoft buy out the rest of open Ai? 0:08:00.960 --> 0:08:03.600 Can Microsoft buy out the rest of open Ai? 0:08:05.200 --> 0:08:06.120 I don't know if they can. 0:08:06.160 --> 0:08:08.480 They obviously have an eleven billion dollar plus stake, so 0:08:08.520 --> 0:08:09.760 they care for the future. 0:08:09.800 --> 0:08:11.200 They want to ensure this happens. 0:08:11.600 --> 0:08:16.400 My view is actually open Ai will get eventively. 0:08:15.880 --> 0:08:17.080 That board settled down. 0:08:17.200 --> 0:08:19.440 We'll get the employee settled down and hopefully keep everyone 0:08:20.040 --> 0:08:23.480 in place. And I don't know their incentive structure, but 0:08:23.560 --> 0:08:25.240 I'm not sure that I'd want to lose all my 0:08:25.400 --> 0:08:28.840 incentive comp and stock options in open Ai just because 0:08:28.840 --> 0:08:30.240 I'm upset the CEO left. 0:08:30.920 --> 0:08:32.360 I don't think that makes a lot of sense. 0:08:32.760 --> 0:08:35.560 So my sense is this ship will be calmed at 0:08:35.600 --> 0:08:39.319 open Ai Microsoft with Via Nadella. 0:08:39.400 --> 0:08:40.960 He's the best negotiator intact. 0:08:41.000 --> 0:08:44.960 He'll figure out a way to ensure that they keep 0:08:44.960 --> 0:08:48.960 this intact, keep their investment preserved, but also have the 0:08:49.000 --> 0:08:52.320 best of having you know, Sam and team there. So 0:08:52.360 --> 0:08:54.920 I look at I think there's going to be a 0:08:54.920 --> 0:08:58.240 better outcome, hopefully for open Ai and their employees, where 0:08:58.520 --> 0:09:01.920 they can get SOMEDISI board members in and get them 0:09:01.960 --> 0:09:05.280 calmed down. I understand why they're is upset, but again, 0:09:05.640 --> 0:09:08.600 I can't recall a story in twenty years where that 0:09:08.679 --> 0:09:12.280 many people signed a petition would walk out. I think 0:09:12.320 --> 0:09:15.200 that you're going to end up seeing a situation where 0:09:15.600 --> 0:09:18.960 there's a middle ground that's met and that hopefully both 0:09:19.000 --> 0:09:22.160 camps survive peacefully and we move on and we can 0:09:22.240 --> 0:09:23.679 of enjoy our Thanksgiving weak. 0:09:23.920 --> 0:09:28.680 Do you think Microsoft remains the key player in AI? Therefore, 0:09:28.920 --> 0:09:33.040 can they remain basically still having the distance between themselves 0:09:33.040 --> 0:09:36.319 and ultimately the viewpoint that alphabet's behind. 0:09:36.320 --> 0:09:38.800 For example, Microsoft's crushing it. 0:09:38.840 --> 0:09:41.240 I was up at the Microsoft Developer Conference Seattle last 0:09:41.280 --> 0:09:43.840 week and you could barely get in to see some 0:09:43.840 --> 0:09:46.400 of these demos. The pace of innovation is like nothing 0:09:46.440 --> 0:09:48.839 I've seen covering tech for twenty years. You talk to 0:09:48.920 --> 0:09:52.600 product managers that are not spokespeople for Microsoft and talk 0:09:52.679 --> 0:09:54.800 about the pace of the innovation. And we were talking 0:09:54.840 --> 0:09:57.160 about months when they said it was go time to 0:09:57.200 --> 0:09:59.800 launch in a product. Now the product isn't perfect, it's 0:09:59.840 --> 0:10:02.640 not gonna it's not going to be the smoothest ride 0:10:02.640 --> 0:10:05.320 out of the gate. But the level of innovation is 0:10:05.360 --> 0:10:07.920 beyond anything we have seen at any any time I've 0:10:07.960 --> 0:10:11.120 spent covering Microsoft, even as a developer of Microsoft in 0:10:11.160 --> 0:10:14.280 the nineties. This pace of innovation is off the charts. 0:10:14.679 --> 0:10:17.480 They are so far ahead of everyone else. And we 0:10:17.520 --> 0:10:21.440 do believe that, you know, Amazon is gonna also be 0:10:21.559 --> 0:10:24.800 in the mix. So our thesis is Amazon and Microsoft. 0:10:24.800 --> 0:10:27.959 When the enterprise Google wins the consumer market, they're very 0:10:27.960 --> 0:10:31.080 different markets, but all three will have a seat at 0:10:31.080 --> 0:10:31.880 the table. 0:10:33.320 --> 0:10:35.600 All right, Jeffrey, Senior Internet analyst Brent Hill on the 0:10:35.640 --> 0:10:38.120 Microsoft side of this story. Coming up on the program, 0:10:38.160 --> 0:10:41.000 we will continue the conversation on Sam Mountain and on 0:10:41.120 --> 0:10:44.440 open ais one. Hundreds of employees threaten to leave the 0:10:44.480 --> 0:10:47.120 company if the board doesn't reside the VC The AI 0:10:47.160 --> 0:11:02.240 reaction is next. This is Bloomberg Technology to recap our 0:11:02.280 --> 0:11:04.640 top story. In a letter to the open ai board 0:11:04.760 --> 0:11:08.360 from hundreds of employees, the vast majority of those left 0:11:08.559 --> 0:11:10.920 that open ai is saying they will resign from the 0:11:11.000 --> 0:11:15.120 company and join the newly announced Microsoft subsidiary led by 0:11:15.120 --> 0:11:17.880 Sam Outman and Greg Brockman. If the board of open 0:11:17.920 --> 0:11:21.280 ai does not resign, lots of implications joining us now, 0:11:21.320 --> 0:11:24.800 as someone that knows open Ai very well worked with 0:11:24.960 --> 0:11:28.640 that company now leads his own startup, Octane. Let's bring 0:11:28.720 --> 0:11:33.000 a match licked Octane AI CEO. First off, what have 0:11:33.040 --> 0:11:35.679 you heard from open ai? You are essentially a developer 0:11:35.679 --> 0:11:39.160 that relies on the underlying technology. You're trying to build 0:11:39.520 --> 0:11:42.840 using underlying technology. Have you had any update from open Ai? 0:11:43.880 --> 0:11:46.640 Yeah, so, Octer the I We've been around since twenty sixteen, 0:11:46.960 --> 0:11:51.200 we now are built on top of open EI's GPT four. 0:11:51.760 --> 0:11:55.040 We help entrepreneurs run AI automations. 0:11:55.120 --> 0:11:58.120 We've helped them generate over half a billion dollars and 0:11:58.160 --> 0:12:00.400 a lot of that is because of what OPENINGI has 0:12:00.480 --> 0:12:05.079 provided us. So this weekend has been really tumultuous. I'm 0:12:05.120 --> 0:12:08.040 a big believer in Sam. The community is a big 0:12:08.080 --> 0:12:10.240 believer in Sam. I know him from back when I 0:12:10.280 --> 0:12:12.280 was my combinator in twenty twelve. 0:12:13.360 --> 0:12:15.440 But open aie we just had dev day. 0:12:15.720 --> 0:12:19.839 I was there. It was the peak of their technology. 0:12:19.960 --> 0:12:23.680 Excitement was off the church. It went really really well. 0:12:24.240 --> 0:12:27.600 Now there's a little bit of a lot of mistrust 0:12:27.679 --> 0:12:28.840 on what's going to happen. 0:12:29.360 --> 0:12:30.880 I am hearing from open. 0:12:30.679 --> 0:12:35.720 Ai that they are expressing support for developers. They're going 0:12:35.800 --> 0:12:37.959 to make sure the API stay up, the technology is 0:12:38.000 --> 0:12:40.800 not going to go down. But this weekend has been 0:12:41.360 --> 0:12:44.040 a lot for anyone in the AI ecosystem. 0:12:44.520 --> 0:12:47.559 Now, let's stay with your relationship with Sam and the 0:12:47.600 --> 0:12:50.160 company from minute. You know lots of people that work there, 0:12:50.640 --> 0:12:53.800 You're friends with people who rely on the ecosystem as well. 0:12:54.640 --> 0:12:57.440 In your minds, is there a path forward where Sam 0:12:57.520 --> 0:13:00.800 comes back to open ai or were looking at a 0:13:00.840 --> 0:13:03.760 new look open ai and trying to understand how you 0:13:03.840 --> 0:13:06.040 work with them and how others work with this new 0:13:06.040 --> 0:13:06.719 look open Ai. 0:13:08.080 --> 0:13:10.760 It's it's really unclear. I think over the weekend we 0:13:10.880 --> 0:13:14.640 thought Sam was going to come back. Now that Satya has, 0:13:14.760 --> 0:13:19.120 you know, announced that he's over at Microsoft, Soya also, 0:13:19.240 --> 0:13:22.480 you know, obviously has to make sure that the Microsoft 0:13:22.520 --> 0:13:24.600 stock is performing well. I don't know if he can 0:13:24.640 --> 0:13:28.000 flip back on that. So it's a really sticky situation. 0:13:28.559 --> 0:13:31.240 It's it's unclear. I have no idea what's going on. 0:13:31.320 --> 0:13:34.120 I don't think that open i employees know what's going. 0:13:34.000 --> 0:13:38.440 On, Matt. Much of this is about a concern of 0:13:38.480 --> 0:13:41.120 the pace of innovation. How you ever had any concern 0:13:41.280 --> 0:13:43.680 are indeed the ability for open ai to live up 0:13:43.679 --> 0:13:45.520 to the amount of demand that we were seeing as well? 0:13:46.880 --> 0:13:50.120 Open ai has treated something that literally is magic. They're 0:13:50.160 --> 0:13:53.240 on the forefront of AI. They're the soul of no 0:13:53.280 --> 0:13:53.600 matter what. 0:13:53.600 --> 0:13:55.959 Anybody wants to say, They're the soul of the AI 0:13:56.040 --> 0:13:59.400 community or the movement. Right now in terms of you know, 0:13:59.440 --> 0:14:01.920 have I thought about is ai going to go too 0:14:02.000 --> 0:14:03.640 fast or is it going to go too slow? 0:14:03.880 --> 0:14:06.720 I think that this disruption is definitely causing it to 0:14:06.760 --> 0:14:09.880 go slower, and I would like it to go faster 0:14:10.120 --> 0:14:12.880 as well as a lot of people in the AI community. 0:14:13.040 --> 0:14:16.160 We rely on this technology. It's allowing you to put 0:14:16.280 --> 0:14:19.880 human like intelligence inside of your code, which means that 0:14:19.920 --> 0:14:24.000 the program can write content, it can talk like a human, 0:14:24.280 --> 0:14:27.040 and it can make reasoning decisions like a human. It 0:14:27.040 --> 0:14:30.720 can be autonomous, and it's just getting better at a 0:14:30.760 --> 0:14:34.880 breakneck speed. So this disruption with Opening Eye, for myself 0:14:35.080 --> 0:14:38.600 and for anyone building on top of this technology, means 0:14:38.600 --> 0:14:41.120 that it isn't going to go slower. Is it going 0:14:41.160 --> 0:14:44.520 to be pulled back? What's going to happen next? It's 0:14:44.840 --> 0:14:48.000 this is not an ideal situation by any means. 0:14:48.280 --> 0:14:50.080 What do you think it means in terms of your 0:14:50.400 --> 0:14:53.480 desire to work with a newly formed unit with a 0:14:53.560 --> 0:14:55.520 Microsoft or indeed without the competitors MAP. 0:14:56.680 --> 0:14:59.680 I think the Opening Eye was seen as someone that 0:14:59.720 --> 0:15:00.720 you could beats. 0:15:00.840 --> 0:15:01.920 It was too big to beats. 0:15:01.920 --> 0:15:05.120 And one of the interesting things that is happening because 0:15:05.120 --> 0:15:08.440 of this weekend, because of all of this chaos, is 0:15:08.760 --> 0:15:11.600 now no matter what happens, let's say they bring back SAM, 0:15:11.680 --> 0:15:16.080 everyone everybody stays open. AI is not infallible. They are 0:15:16.120 --> 0:15:18.800 someone you can beat now, and that actually opens up 0:15:18.800 --> 0:15:22.280 a tremendous amount of opportunities for new people, whether they're 0:15:22.320 --> 0:15:26.120 ex employees of open ai or other entrepreneurs, to go 0:15:26.240 --> 0:15:29.040 start companies that compete with open ai. And I think 0:15:29.080 --> 0:15:31.560 that that's going to happen. This leaves a huge open 0:15:31.720 --> 0:15:33.720 for that. Any event that a lot of people leave 0:15:33.760 --> 0:15:36.360 open Ai, maybe a bunch joined Microsoft and go join 0:15:36.440 --> 0:15:38.360 Sam and Greg, a lot of the other ones are 0:15:38.360 --> 0:15:41.280 going to start companies, and I think more capital than 0:15:41.400 --> 0:15:43.920 ever is going to flow into AI. 0:15:44.680 --> 0:15:46.200 You know, just this year, I. 0:15:46.240 --> 0:15:49.200 Started a venture capital firm called Theory Forge Ventures with 0:15:49.240 --> 0:15:50.640 Ben Park to do just that. 0:15:51.040 --> 0:15:53.080 And this is a sad situation with open ai. 0:15:53.240 --> 0:15:56.520 It makes me sad, but at the same time investors 0:15:56.560 --> 0:15:57.840 are going to be clamoring in. 0:15:58.320 --> 0:16:01.760 Mat Schicked obtain Ai CEO and indeed avention Capitalist. We 0:16:01.800 --> 0:16:10.440 thank you so much. Turning back to where we left 0:16:10.440 --> 0:16:13.240 off with our open Ai conversation, the drama, the impact 0:16:13.240 --> 0:16:16.480 on the AI landscape more broadly, Sarahgayo, founder of Conviction. 0:16:16.560 --> 0:16:19.600 It's a venture firm purpose built to serve AI native 0:16:19.640 --> 0:16:21.880 companies and what we were just talking to Matchlick who 0:16:21.920 --> 0:16:24.440 thinks that more innovation will start ups are going to 0:16:24.480 --> 0:16:27.240 come of this, Sarah, do you abide by that silver 0:16:27.320 --> 0:16:29.000 lining to what feels like a cloud right now? 0:16:31.920 --> 0:16:33.680 I mean I would start by saying I have such 0:16:33.800 --> 0:16:36.120 enormous respect for the work Sam and Greg and Illi 0:16:36.160 --> 0:16:38.520 and the entire openI team have done over the years. 0:16:38.760 --> 0:16:41.400 Right So before we talk about silver aligning, I think 0:16:41.400 --> 0:16:44.480 we should talk about the contribution they've had in terms 0:16:44.480 --> 0:16:47.080 of igniting a technology revolution. It's just very sad to 0:16:47.120 --> 0:16:49.520 see it happen. And since your respect to Ilia for 0:16:50.200 --> 0:16:54.040 you know, a public apology for something he regrets, whatever 0:16:54.040 --> 0:16:56.240 the outcome, this is his life's for work and he 0:16:56.760 --> 0:16:59.600 has made a tremendous contribution to the enhancement of AI 0:17:00.160 --> 0:17:02.800 and many of the startups that we back and partner 0:17:02.800 --> 0:17:06.960 with open ai in like Harvey, which is legal AI 0:17:07.359 --> 0:17:10.160 and a portfolio company for open Ai. We benefit from 0:17:10.200 --> 0:17:15.959 all of that innovation. I think that being said, the ecosystem, 0:17:16.160 --> 0:17:20.960 the landscape is much more open than it was a 0:17:21.000 --> 0:17:23.960 week ago, which is wild to say. Opening is both 0:17:24.040 --> 0:17:27.359 revered and feared. In its current state, it is much diminished. 0:17:28.960 --> 0:17:31.480 Sarah, delight to have you on the show. You know 0:17:31.600 --> 0:17:34.960 everyone you know the players. You yourself and not invested 0:17:35.000 --> 0:17:39.000 in open ai, but the ecosystem you are so closely 0:17:39.440 --> 0:17:46.760 connected to. Just tell me your reaction, frankly, your how 0:17:46.800 --> 0:17:48.960 you felt this weekend watching that unfold. 0:17:50.560 --> 0:17:53.960 Yeah, I mean I started with sort of respect and 0:17:54.040 --> 0:17:59.080 sadness for friends who work at open AI. I think 0:17:59.160 --> 0:18:03.520 my secondary action is really just as champion level business people, 0:18:03.560 --> 0:18:05.679 you have to hand it to Saya and Kevin and 0:18:05.760 --> 0:18:09.720 managing all of this. The multiple on Microsoft stocks and 0:18:09.840 --> 0:18:11.760 Saya took it over in twenty fourteen is up ten 0:18:11.960 --> 0:18:14.360 x from what was already one of the largest five. 0:18:14.200 --> 0:18:14.960 Companies in the world. 0:18:15.040 --> 0:18:21.360 So I am and I'm just floored by their ability 0:18:21.359 --> 0:18:24.399 to navigate the situation. It's incredibly hard to be this 0:18:24.560 --> 0:18:29.240 aggressive and strategic and steering so large a ship and 0:18:29.320 --> 0:18:31.679 anticipating the title wave and coming out so far ahead. 0:18:31.760 --> 0:18:35.720 So all roads from here are paved the gold for Microsoft. 0:18:36.359 --> 0:18:39.440 I think, going back to the first question, we are 0:18:39.480 --> 0:18:43.679 an AI focused early stage investing firm, right and so 0:18:44.359 --> 0:18:46.640 a part of it is a recognition that there are 0:18:46.720 --> 0:18:50.879 opportunities at every layer, which we continue to believe, but 0:18:51.000 --> 0:18:54.840 now even more so for the independent foundation model labs 0:18:54.840 --> 0:18:59.240 such as mistral and inflection and anthropic, and I think 0:18:59.240 --> 0:19:04.600 there I think there's increasing interest in open source from yes, 0:19:05.000 --> 0:19:08.919 corporations and developers just in having more control over their 0:19:08.920 --> 0:19:12.000 own models and running them monitor on their own infrastructure, 0:19:12.040 --> 0:19:14.560 and infrint platforms like Base ten, and so there will 0:19:14.600 --> 0:19:19.000 be continued capital for startups making these bets from investors 0:19:19.000 --> 0:19:21.720 like in Conviction, and of course we're full stack and 0:19:21.760 --> 0:19:23.400 application level companies as well. 0:19:24.200 --> 0:19:27.639 You've invested in startups building your own large language models. 0:19:27.640 --> 0:19:30.760 You've interviewed on your podcast leaders some of the firms 0:19:30.800 --> 0:19:34.760 with making bigger models with much many more billions of parameters. 0:19:35.160 --> 0:19:39.679 Does this weekend make it a more competitive marketplace for 0:19:39.760 --> 0:19:42.400 building generative AI technology? 0:19:45.040 --> 0:19:49.360 I think companies are. So there's sort of two dimensions, right, 0:19:49.400 --> 0:19:54.000 There's like what it takes to compete, and of course 0:19:54.040 --> 0:19:58.280 like the reaction from the ecosystem of customers and developers. 0:19:58.880 --> 0:20:02.080 This set of events, of course gives all of open 0:20:02.080 --> 0:20:06.720 AI's customers and developers pause. They're making long term bets 0:20:06.760 --> 0:20:09.080 and they want to understand the incentives and staying power 0:20:09.160 --> 0:20:13.920 of their partners. My experienced startup founders are sometimes reluctant 0:20:13.960 --> 0:20:16.959 to ask for the early check or seem too profit 0:20:17.000 --> 0:20:20.640 seeking or capitalists to their customers. But I think transparency 0:20:20.720 --> 0:20:22.920 is key. That's a lesson from this weekend. I've heard 0:20:23.040 --> 0:20:25.879 so many customers ask startups, how are you planning to 0:20:25.880 --> 0:20:28.040 make money? Or even ask me as an investor in 0:20:28.080 --> 0:20:30.520 these starts, is the business viable? I am willing to pay, 0:20:30.560 --> 0:20:32.800 but I don't want to make the wrong bet. Well, 0:20:32.840 --> 0:20:36.960 they make it, and the nonprofit board here is explicitly 0:20:37.080 --> 0:20:41.080 prioritizing other goals they have above customers and developers, and 0:20:41.080 --> 0:20:44.280 so I think that I think that does change the 0:20:44.359 --> 0:20:49.959 ecosystem in terms of other opportunities, like we should not 0:20:50.040 --> 0:20:53.240 trivialize the work that open Ai had done has done 0:20:53.280 --> 0:20:57.240 over the years. Right, creating foundation models is a huge effort, 0:20:57.320 --> 0:21:07.040 and so from an infrastructure, data, partnerships, pooling and talent perspective, 0:21:07.960 --> 0:21:11.639 I think you know the ship of theseus. It's a 0:21:11.640 --> 0:21:14.119 thought experiment, like if you rebuild a ship entirely with 0:21:14.200 --> 0:21:16.920 new parts, is it the same ship. It's a question 0:21:16.960 --> 0:21:19.720 of identity, and it's often applied to human consciousness and 0:21:19.800 --> 0:21:22.080 sci fi, but it applies to organizations too. 0:21:22.200 --> 0:21:24.320 Right, if okai has the. 0:21:24.320 --> 0:21:26.560 Weights and then all of the advantages that we were 0:21:26.600 --> 0:21:30.280 just talking about. They're huge, but an organization is also 0:21:30.320 --> 0:21:31.320 a sera culture. 0:21:31.760 --> 0:21:34.840 Yeah, yeah, conviction fanta serago. It's just great to have 0:21:34.880 --> 0:21:36.840 you on the show and get that perspective. Thank you. 0:21:44.400 --> 0:21:46.680 Some breaking news for you right now the United States, 0:21:46.760 --> 0:21:49.200 the Justice Department, being exact, is seeking more than four 0:21:49.240 --> 0:21:52.680 billion dollars from Finance, that of course being the cryptocurrency 0:21:52.720 --> 0:21:56.800 exchange to end a criminal case. It's years long investigation 0:21:57.160 --> 0:21:59.560