1 00:00:00,040 --> 00:00:03,640 Artificial intelligence. AI AI AI AI AI AI AI. 2 00:00:04,280 --> 00:00:08,160 AI 2025 has been another crazy year for artificial 3 00:00:08,160 --> 00:00:10,480 intelligence. We've seen the rise of thinking 4 00:00:10,480 --> 00:00:13,080 models. We've seen video creation, 5 00:00:13,320 --> 00:00:16,800 including Open AI's fictionalized version of it. 6 00:00:16,920 --> 00:00:20,320 We are about to host the Cerebral Valley AI Summit, our 7 00:00:20,360 --> 00:00:24,880 ultimate insider AI event for startup founders and investors. 8 00:00:25,160 --> 00:00:28,800 I host that event with Max Child and James Wilsterman, my dear 9 00:00:28,800 --> 00:00:31,840 friends and the cofounders of the voice AI games company 10 00:00:31,840 --> 00:00:33,840 Volley. They're coming on the show to 11 00:00:33,840 --> 00:00:37,040 pick up where we left off on our AI startup draft. 12 00:00:37,120 --> 00:00:41,040 We've been picking the biggest names in AI startups, and it's a 13 00:00:41,040 --> 00:00:44,040 good way to learn about the top Silicon Valley startups that 14 00:00:44,040 --> 00:00:45,400 insiders are paying attention to. 15 00:00:45,400 --> 00:00:48,520 So we're going to have a fun time picking up new companies 16 00:00:48,520 --> 00:00:51,920 and dropping them from our list. But first we will discuss some 17 00:00:51,920 --> 00:00:55,400 of the key themes in AI going into Seruville Valley Volley 18 00:00:55,520 --> 00:00:58,560 thinking through what's really happened over the last 12 19 00:00:58,560 --> 00:00:59,720 months. So give it a listen. 20 00:01:07,920 --> 00:01:12,200 All right, let's get into it. We've got the Srebul Valley AI 21 00:01:12,200 --> 00:01:16,440 Summit coming up November 12th. I'm here with Maxwell and James 22 00:01:16,440 --> 00:01:20,280 Wilsterman, my Srebul Valley Co host, good friends and the Co 23 00:01:20,280 --> 00:01:24,120 founders of the voice AI games company Volley. 24 00:01:25,160 --> 00:01:29,960 We have two main objectives for this episode 1 themes, which is 25 00:01:30,200 --> 00:01:33,720 half preparation for the event for ourselves and half just sort 26 00:01:33,720 --> 00:01:37,960 of like, OK, where where are we in sort of the zeitgeist of AI 27 00:01:37,960 --> 00:01:41,440 going into this event? And then we'll see what comes 28 00:01:41,440 --> 00:01:45,720 out of it afterwards. And then to our favorite part of 29 00:01:45,720 --> 00:01:49,640 the show, we're going to update our our drafts in a second. 30 00:01:49,640 --> 00:01:51,040 So hey guys, welcome to the show. 31 00:01:51,440 --> 00:01:52,680 Hello, Eric. Glad to be here. 32 00:01:52,920 --> 00:01:54,520 Hey, Eric, hope you're doing well. 33 00:01:55,080 --> 00:01:57,120 Yeah, I've got a screaming child. 34 00:01:58,240 --> 00:01:59,920 We're a day. We're 28. 35 00:01:59,920 --> 00:02:02,040 Days the draft gets. Heated you get in. 36 00:02:02,280 --> 00:02:04,640 In trouble over there. Yeah, keep it down. 37 00:02:04,640 --> 00:02:07,320 No, I have enough distance that I think I'm far away, but a 38 00:02:07,320 --> 00:02:10,160 screaming child can can carry quite far. 39 00:02:11,760 --> 00:02:13,560 Great. And you guys have been dialed in 40 00:02:13,960 --> 00:02:16,040 thinking about sort of the big theme. 41 00:02:16,040 --> 00:02:21,360 So let's let's get into it. Yeah, I think the goal here is 42 00:02:21,360 --> 00:02:24,880 everybody come with at least two, if not sort of three themes 43 00:02:24,880 --> 00:02:29,880 that you think will stand out in AI generative AII mean we do 44 00:02:30,640 --> 00:02:34,200 Cerebral Valley twice a year. We did London mid year, we do 45 00:02:34,360 --> 00:02:38,600 San Francisco in November. So these are themes, you know, 46 00:02:38,600 --> 00:02:41,440 meant to, I think look at like a six month window or sort of 47 00:02:41,440 --> 00:02:46,120 what's, what's the mood for the next 6 months in AI Max? 48 00:02:46,120 --> 00:02:47,520 I know you've thought about this a lot you. 49 00:02:47,520 --> 00:02:49,720 Want to go first. I mean, I have a bit, I have a 50 00:02:49,720 --> 00:02:53,120 bit of a look back and a look forward here in terms of 2025 51 00:02:53,120 --> 00:02:55,320 and beyond. I think that what's interesting 52 00:02:55,320 --> 00:02:58,440 about this year is there were a lot of kind of industry 53 00:02:58,440 --> 00:03:02,640 prognosticators trying to call certain themes for 2025 back in 54 00:03:02,640 --> 00:03:07,120 like January, February, March. I remember distinctly I had some 55 00:03:07,120 --> 00:03:09,840 very intimate conversations with folks who worked at Open AI. 56 00:03:10,360 --> 00:03:11,240 Who? Intimate. 57 00:03:11,720 --> 00:03:13,160 Intimate they. Were intimate. 58 00:03:13,280 --> 00:03:14,480 You're like in Asano. Yeah, yeah. 59 00:03:16,080 --> 00:03:18,320 Yeah, Yeah. Let me tell you. 60 00:03:19,080 --> 00:03:23,320 The Gary Chan sauna. A couple cocktails, sauna 61 00:03:23,320 --> 00:03:24,600 action. It was a great time. 62 00:03:26,040 --> 00:03:29,320 I won't say whose house it was at anyway. 63 00:03:29,920 --> 00:03:31,560 Yeah. And they basically were like, 64 00:03:31,800 --> 00:03:34,160 there's going to be two big themes for 2025. 65 00:03:34,840 --> 00:03:37,320 And I don't know if you guys want to guess these in advance. 66 00:03:37,320 --> 00:03:40,360 They were like the two big themes are 2025 is the year of 67 00:03:40,400 --> 00:03:45,320 agents and 2025 is the year of voice. 68 00:03:45,960 --> 00:03:49,400 And my contention would be that both of these themes turned out 69 00:03:49,400 --> 00:03:54,440 to be bullshit right too early. I mean, like, I think that we 70 00:03:54,440 --> 00:03:56,200 talked about this a year ago on the podcast. 71 00:03:56,440 --> 00:03:58,920 I don't even think we'd seen a quote UN quote agent at that 72 00:03:58,920 --> 00:04:00,960 point. And I think at this point you 73 00:04:00,960 --> 00:04:03,160 could argue maybe we're starting to see the contours of what 74 00:04:03,160 --> 00:04:05,800 people would define as an agent, but it's almost exclusively 75 00:04:05,800 --> 00:04:10,240 encoding and then voice. Look, James and I are obviously 76 00:04:10,440 --> 00:04:13,200 voice bandwagon ears. You know, we may have hand 77 00:04:13,200 --> 00:04:17,640 constructed the voice bandwagon out of Wood and Nails about 8 78 00:04:17,640 --> 00:04:21,000 years ago, but I don't really think 22 in the voice either. 79 00:04:21,000 --> 00:04:25,720 I think there's been progression in voice recognition in TTS, 11 80 00:04:25,720 --> 00:04:27,600 labs, evaluations, things like that. 81 00:04:27,600 --> 00:04:29,760 But I certainly wouldn't say we're all spending every day 82 00:04:29,760 --> 00:04:33,440 talking of voice AI and even the products that we use I don't 83 00:04:33,440 --> 00:04:35,360 think have made maybe as much progress as we would have 84 00:04:35,360 --> 00:04:38,000 expected. So that was my initial framing 85 00:04:38,000 --> 00:04:40,240 for this, which is those were the two expected themes for. 86 00:04:40,240 --> 00:04:42,680 The what didn't happen? Yeah, it didn't happen, James. 87 00:04:42,680 --> 00:04:44,840 Your reaction to the ones. That did happen, but yeah, go 88 00:04:44,840 --> 00:04:45,240 ahead. Yeah. 89 00:04:45,600 --> 00:04:49,720 I mean, I I agree, I think, I think that a lot of companies 90 00:04:49,720 --> 00:04:53,360 built out foundations for agents. 91 00:04:53,360 --> 00:04:58,800 I think Open AI just came out with their kind of agent's drag 92 00:04:58,800 --> 00:05:01,000 and drop framework and there have been a lot of other 93 00:05:01,000 --> 00:05:06,360 companies entering that space. There's like an agent's SDK from 94 00:05:06,360 --> 00:05:09,440 Open AI and there's other companies that have these now. 95 00:05:09,440 --> 00:05:16,280 But I agree, I haven't seen like deployment or scale of these in 96 00:05:16,280 --> 00:05:20,480 any meaningful way. I, I think the fact that we're 97 00:05:20,480 --> 00:05:24,760 talking, that people talk about agents is itself a sign that 98 00:05:24,760 --> 00:05:26,160 things aren't where they need to be, right? 99 00:05:26,160 --> 00:05:29,800 Like coding customer support, like if you're succeeding, you 100 00:05:29,800 --> 00:05:31,440 just say, oh, there's this great use case. 101 00:05:31,440 --> 00:05:34,440 And like, I, I feel like you don't start talking about SAS 102 00:05:34,440 --> 00:05:37,520 and software as a service until it's like, OK, you've got some 103 00:05:37,520 --> 00:05:40,960 products people can point to. It's like you vertical SAS 104 00:05:40,960 --> 00:05:43,360 becomes a thing because you've got like mind body and you say, 105 00:05:43,360 --> 00:05:47,120 oh, you can, you know, provide for this sector, We'll do it for 106 00:05:47,120 --> 00:05:49,400 everybody else. But I feel like with agents and 107 00:05:50,440 --> 00:05:52,480 you know, this isn't the first time this ever happened in 108 00:05:52,480 --> 00:05:56,840 technology, but people are like, you know, sort of in concluding 109 00:05:56,840 --> 00:05:59,760 how all this can play out without sort of the first really 110 00:05:59,760 --> 00:06:04,280 successful use of agents. Because we say coding, but even 111 00:06:04,280 --> 00:06:07,960 with coding, it's like a lot of it is a sidekick, right? 112 00:06:07,960 --> 00:06:11,600 And not necessarily like, you know, Devon running, running on 113 00:06:11,600 --> 00:06:12,800 its own. Yeah. 114 00:06:13,200 --> 00:06:16,200 I mean, I think that Andre Carpathy obviously had a very 115 00:06:16,200 --> 00:06:19,040 notable podcast a couple weeks ago on Dwarkish podcast where he 116 00:06:19,040 --> 00:06:20,400 talked about development and agents. 117 00:06:20,400 --> 00:06:24,000 And I think the analogy he used was to self driving cars, which 118 00:06:24,000 --> 00:06:26,640 I thought was a good one where but more like self driving cars 119 00:06:26,640 --> 00:06:30,160 back in 2016 or 2017 where he's saying, OK, if I think of an 120 00:06:30,160 --> 00:06:33,520 agent as essentially a smart intern who can go off and do 121 00:06:33,520 --> 00:06:35,680 basically any sort of task that I assigned them. 122 00:06:36,040 --> 00:06:39,280 Like, you know, it can do a little bit of stuff and you give 123 00:06:39,280 --> 00:06:43,120 it a very specific use case and a very focused set of, you know, 124 00:06:43,120 --> 00:06:44,720 check boxes it needs to check off. 125 00:06:44,920 --> 00:06:46,960 It can occasionally go off and do interesting things. 126 00:06:47,400 --> 00:06:50,600 But you know, his point was what happened in self driving cars is 127 00:06:50,600 --> 00:06:53,720 we, we, we ended up having to go through what he called like the 128 00:06:53,720 --> 00:06:57,240 March of the March of nines, the March of nines being what 129 00:06:57,240 --> 00:06:59,360 percentage accurate is this thing right? 130 00:06:59,360 --> 00:07:03,000 And we started with 90% and then we got to 99% and then we got to 131 00:07:03,000 --> 00:07:06,440 99.9% and then 99.99% and so on and so forth. 132 00:07:06,440 --> 00:07:09,960 And ultimately to be superior to a human, you know, Waymo, you 133 00:07:09,960 --> 00:07:12,920 maybe need 5 nines or six nines or 7 nines or whatever, right? 134 00:07:13,320 --> 00:07:16,960 And he's saying, hey, agents are maybe at the first nine, you 135 00:07:16,960 --> 00:07:19,320 know, today, maybe this, you know, you could argue the second 136 00:07:19,320 --> 00:07:21,240 time, but I don't even think you could argue that really. 137 00:07:21,600 --> 00:07:25,800 And we're probably still 234 more kind of nines away and and 138 00:07:25,800 --> 00:07:27,920 maybe things are getting 10 times better every year. 139 00:07:27,920 --> 00:07:30,880 So that's only three or four or five years, but he put more like 140 00:07:30,880 --> 00:07:34,360 a 10 year timeline on sort of fully, you know, human capable 141 00:07:34,360 --> 00:07:37,520 level agents. Why do you think companies like 142 00:07:37,520 --> 00:07:43,120 Open AI are moving so fast in this direction? 143 00:07:43,120 --> 00:07:47,440 Building out the SDK, is that the the no code interfaces the 144 00:07:48,080 --> 00:07:52,080 browser agents within their, their their Atlas browser. 145 00:07:53,720 --> 00:07:57,360 I guess like why? What is it marketing or is is 146 00:07:57,360 --> 00:07:59,560 there a belief that this is further alignment? 147 00:07:59,560 --> 00:08:03,480 It is or or you know, a different take on how the 148 00:08:03,480 --> 00:08:06,120 timelines might look. I mean, my argument would be 149 00:08:06,120 --> 00:08:07,760 it's competitive defense, basically. 150 00:08:07,760 --> 00:08:10,280 Like there's people doing these little agentic drag and drop 151 00:08:10,280 --> 00:08:11,720 frameworks and they're like, we better do that. 152 00:08:12,000 --> 00:08:14,920 And then there's people doing, you know, browsers and they're 153 00:08:14,920 --> 00:08:16,240 like, maybe the browsers will be a thing. 154 00:08:16,240 --> 00:08:17,600 I guess we better fucking do a browser. 155 00:08:17,600 --> 00:08:20,240 You know, like it's like, that's my argument. 156 00:08:20,240 --> 00:08:22,360 I could be wrong. You know, obviously they're much 157 00:08:22,360 --> 00:08:24,400 smarter than we are. They're smarter than I am. 158 00:08:24,800 --> 00:08:26,800 Like they figured out a lot of stuff that we would never 159 00:08:26,800 --> 00:08:28,720 figured out. So maybe they see the future, 160 00:08:28,720 --> 00:08:31,480 they see the 2-3 year agentic vision and it's it's, you know, 161 00:08:31,480 --> 00:08:32,960 they're building towards that right now. 162 00:08:33,200 --> 00:08:34,600 I think. Yeah, maybe the more charitable 163 00:08:34,600 --> 00:08:37,480 interpretation is they can see where this thing's going to be 164 00:08:37,480 --> 00:08:39,520 in three or five years and that we're putting all the 165 00:08:39,520 --> 00:08:43,039 foundations in place today, but it doesn't seem like they work 166 00:08:43,039 --> 00:08:44,920 that well. Like I have not heard of anyone 167 00:08:44,920 --> 00:08:48,320 using this agent drag and drop composer thing for to build an 168 00:08:48,320 --> 00:08:50,600 agent in the last four weeks since they launched it. 169 00:08:50,600 --> 00:08:52,360 Have you? I mean, I'm I'm it like 170 00:08:52,360 --> 00:08:55,640 basically vanished without a whisper in in my my social 171 00:08:55,640 --> 00:08:59,680 feeds, I guess. I'm going to offer a theme, I 172 00:08:59,680 --> 00:09:04,680 guess in it feels like it it follows the same pitfalls of the 173 00:09:04,680 --> 00:09:06,280 ones that you said didn't come true. 174 00:09:06,280 --> 00:09:09,280 It's sort of like it's you know, it's going to be voice, it's 175 00:09:09,280 --> 00:09:12,720 going to be agents. And I'm going to say like 176 00:09:13,120 --> 00:09:17,680 fiction frenzy, you know, it's with the Sora creation and just, 177 00:09:17,960 --> 00:09:21,440 you know, and, and also speaking to your 999 issue, it's like, 178 00:09:21,440 --> 00:09:27,320 OK, what doesn't require stuff that's like, you know, totally 179 00:09:27,320 --> 00:09:29,520 accurate works well. Like, you know, the problem with 180 00:09:29,520 --> 00:09:31,600 agents is compounding. It's like you, you make a 181 00:09:31,600 --> 00:09:33,520 mistake here. You have even more mistakes down 182 00:09:33,520 --> 00:09:36,640 the line. So just just the, the errors are 183 00:09:36,640 --> 00:09:38,880 are huge. Where stuff where it's like 184 00:09:38,880 --> 00:09:43,760 create, you know, regular people create content, fictionalized 185 00:09:43,760 --> 00:09:47,320 stuff. I think, I think that's has a 186 00:09:47,320 --> 00:09:49,880 lot of promise. And so, and then everybody 187 00:09:49,880 --> 00:09:52,560 copies open AI and so they have their Sora app. 188 00:09:52,560 --> 00:09:55,080 It got a lot of downloads and it's like, oh, that can work. 189 00:09:55,160 --> 00:09:57,480 You know, we should, we should lean into that. 190 00:09:58,200 --> 00:10:01,360 And the last thing I'll offer on this theme, maybe it's an overly 191 00:10:01,360 --> 00:10:05,800 expansive view, but I'm trying to also capture the clearly 192 00:10:05,800 --> 00:10:09,560 serious like mental health thing and the sort of like people 193 00:10:09,560 --> 00:10:13,040 spinning up their own worlds and how that's going to loom in the 194 00:10:13,200 --> 00:10:15,840 AI story too. It's like you've got, it's sort 195 00:10:15,840 --> 00:10:19,840 of, this is sort of my consumer, consumers going wild with what 196 00:10:19,840 --> 00:10:23,640 is possible with AI and who cares how accurate it is at the 197 00:10:23,640 --> 00:10:26,720 moment. Why is that a tale of things 198 00:10:26,720 --> 00:10:28,960 that didn't work out? Or how are you thinking? 199 00:10:29,120 --> 00:10:31,440 Of that just because it's a very like oh, we're going to build a 200 00:10:31,440 --> 00:10:34,920 sore app or you know, it's like you see a glimmer of something 201 00:10:34,920 --> 00:10:37,080 that's like you see coding and so we're all going to have 202 00:10:37,080 --> 00:10:40,800 agents in every category you see you know, maybe a little funding 203 00:10:40,800 --> 00:10:42,360 going to voice. I just, I just think it's the 204 00:10:42,360 --> 00:10:46,520 sort of like you got, you got a sign of some success and then 205 00:10:46,520 --> 00:10:49,160 the AR world is desperate to see more. 206 00:10:49,160 --> 00:10:51,880 And so you're saying, oh, there's going to be so much much 207 00:10:51,880 --> 00:10:54,360 building there. So I think it's just reading a 208 00:10:54,360 --> 00:10:58,040 little bit too much maybe from one area of success. 209 00:10:58,320 --> 00:11:01,160 But I mean, you guys are in the you're trying to do fiction AI 210 00:11:01,160 --> 00:11:02,160 stuff. It's your. 211 00:11:02,160 --> 00:11:03,280 Yeah, it's possible. Yeah. 212 00:11:03,280 --> 00:11:06,880 I mean, I was gonna sort of say, I think that AI video is the one 213 00:11:06,880 --> 00:11:09,360 that kind of came out of nowhere this year as like an actual 214 00:11:09,360 --> 00:11:11,920 theme of the year in some ways. Whether or not you think it's 215 00:11:11,920 --> 00:11:14,400 good enough yet, I think that's an interesting argument. 216 00:11:14,400 --> 00:11:20,920 But I think that VO3 in May and then SORA a month ago definitely 217 00:11:20,920 --> 00:11:24,440 like sort of felt like minor nuclear explosions in this sort 218 00:11:24,440 --> 00:11:27,360 of consumer scene, right? Especially the especially the 219 00:11:27,360 --> 00:11:32,640 Sora era, where you could RIP off classic copyrighted IPS. 220 00:11:32,680 --> 00:11:35,480 Yeah, for like a week or five days or whatever it was. 221 00:11:35,840 --> 00:11:40,280 Make a SpongeBob episode of yourself and any character from 222 00:11:40,280 --> 00:11:44,280 history you wanted. And I just felt like that those 223 00:11:44,280 --> 00:11:47,840 seem to achieve actual consumer breakouts in a meaningful way. 224 00:11:47,840 --> 00:11:50,960 And and they fell off a big Cliff, I think when they pulled 225 00:11:50,960 --> 00:11:52,600 all the copyrighted material out. 226 00:11:53,120 --> 00:11:56,600 But I do think that AI video is something that has kind of 227 00:11:56,600 --> 00:12:00,440 crossed the chasm into actually useful now, whereas maybe a year 228 00:12:00,440 --> 00:12:02,200 ago it was sort of more of a curiosity. 229 00:12:02,680 --> 00:12:05,760 I think again, you can make the case that it's not a part of our 230 00:12:05,760 --> 00:12:07,960 everyday lives or whatever, but it feels like it made a big 231 00:12:07,960 --> 00:12:11,000 progress. And then I do think the AI 232 00:12:11,000 --> 00:12:13,760 psychosis theme, which is a little more of a negative theme, 233 00:12:14,640 --> 00:12:17,680 is a real 1 and probably a big part of the future of our 234 00:12:17,680 --> 00:12:22,520 lifetimes is this sort of AIAI induced psychosis and AI induced 235 00:12:22,840 --> 00:12:26,560 dream worlds and virtual friends and can they drive people off 236 00:12:26,560 --> 00:12:27,760 the ledge and all this kind of stuff. 237 00:12:27,760 --> 00:12:32,400 So kind of a sad one, but probably a very real part of the 238 00:12:32,400 --> 00:12:36,640 future. And do you guys think that based 239 00:12:36,640 --> 00:12:39,680 on what you're seeing that is happening at large scale like 240 00:12:39,680 --> 00:12:44,160 the AI itself is the cause of psychosis or do you think? 241 00:12:44,160 --> 00:12:46,920 Well, wasn't Sam Altman? Seems like he's owning this 242 00:12:46,920 --> 00:12:48,560 issue A. Little bit, yeah. 243 00:12:49,040 --> 00:12:50,320 Yeah. I mean, I think he was saying 244 00:12:50,320 --> 00:12:52,120 that very recently. Yeah. 245 00:12:52,200 --> 00:12:56,960 I mean, I was trying to read what he was saying, and a lot of 246 00:12:56,960 --> 00:13:01,800 it was kind of couched in this idea that they have massive 247 00:13:01,800 --> 00:13:03,920 scale. They have 800, right? 248 00:13:04,520 --> 00:13:08,880 At scale, all humans problem. Human problems are yeah, open AI 249 00:13:08,880 --> 00:13:12,080 problems like. What percentage of non AI 250 00:13:12,080 --> 00:13:15,320 conversations every day show signs of psychosis in the world? 251 00:13:15,320 --> 00:13:17,080 Right? Right, A lot. 252 00:13:17,080 --> 00:13:19,920 Exactly. It's just, yeah, we just have an 253 00:13:19,920 --> 00:13:22,600 active psych exam. Yeah, now we have someone 254 00:13:22,600 --> 00:13:23,600 running. All the time we. 255 00:13:23,600 --> 00:13:29,000 Can actually monitor this. So I'm just curious to see the 256 00:13:29,000 --> 00:13:33,480 causality evolve. I think the current chat bots 257 00:13:33,720 --> 00:13:36,680 would be net good in that they're trying to ground people 258 00:13:36,680 --> 00:13:39,480 in reality. If anything, I'm concerned about 259 00:13:39,480 --> 00:13:43,760 where we're going, which is this embrace of what the social media 260 00:13:43,760 --> 00:13:48,080 companies did, which is moving. You know, Open AI, Anthropic and 261 00:13:48,080 --> 00:13:50,480 everybody else wanted accurate answers. 262 00:13:50,480 --> 00:13:52,000 You know, that works in the enterprise. 263 00:13:52,000 --> 00:13:53,920 That was sort of what researchers building. 264 00:13:54,280 --> 00:13:56,600 But obviously the social media world learned, oh, we want 265 00:13:56,600 --> 00:13:57,560 engagement. Who cares? 266 00:13:58,320 --> 00:14:00,400 The humans can decide if it's accurate. 267 00:14:00,400 --> 00:14:03,680 And so, yeah, we get things can get worse. 268 00:14:04,120 --> 00:14:07,560 But right now, if the models are trying to be accurate, often to 269 00:14:07,560 --> 00:14:10,560 me, it's bringing me back towards reality. 270 00:14:10,560 --> 00:14:13,320 You know, it's like, OK, no, don't worry too much about that. 271 00:14:13,320 --> 00:14:16,440 This is what's normal. I think they're that's true, but 272 00:14:16,480 --> 00:14:24,720 I think clearly like especially pre GPT 5 and maybe in some 273 00:14:24,720 --> 00:14:27,840 other apps, like you definitely saw models that had that 274 00:14:27,840 --> 00:14:31,280 sycophant C problem that would like, you know, eventually go 275 00:14:31,280 --> 00:14:34,120 off the rails as the conversation and the context 276 00:14:34,120 --> 00:14:38,000 window got longer and just start agreeing with you. 277 00:14:38,800 --> 00:14:43,840 So that I think is the risk vector that still exists, you 278 00:14:43,840 --> 00:14:48,160 know, but even even if the models you know on your first 279 00:14:48,160 --> 00:14:51,320 attempt are pushing back. For anyone running a startup 280 00:14:51,320 --> 00:14:54,640 right now, Mongo DB for Startups is one of those programs that's 281 00:14:54,640 --> 00:14:57,360 looking to make the life of a founder easier. 282 00:14:57,680 --> 00:15:00,680 It's basically a shortcut for founders who want to move faster 283 00:15:00,800 --> 00:15:03,560 and spend less. You get free credits on Mongo DB 284 00:15:03,560 --> 00:15:06,920 Atlas, so your products ready to scale from day one without 285 00:15:06,920 --> 00:15:10,520 worrying about database costs. They also throw in 200 million 286 00:15:10,520 --> 00:15:13,640 Voyage AI tokens so you can start experimenting with AI 287 00:15:13,640 --> 00:15:16,400 features right out of the gate. You'll get one-on-one time with 288 00:15:16,400 --> 00:15:19,360 actual Mongo DB engineers, people who can help you 289 00:15:19,360 --> 00:15:22,520 troubleshoot, optimize, and think through your next stage of 290 00:15:22,520 --> 00:15:24,200 growth. There's even Co marketing and 291 00:15:24,200 --> 00:15:26,920 partner opportunities, which means more visibility and 292 00:15:26,920 --> 00:15:29,400 potential customers without extra spend. 293 00:15:29,800 --> 00:15:32,160 If you're building, scaling, or just trying to stretch your 294 00:15:32,160 --> 00:15:36,240 runway a little longer, Mongo DB for startups can actually help. 295 00:15:36,320 --> 00:15:40,640 Head to mongodb.com/startups to learn more. 296 00:15:41,320 --> 00:15:45,240 I think a theme for the conference this year will be how 297 00:15:45,240 --> 00:15:55,000 much AI is holding up the US economy and whether I think 298 00:15:55,000 --> 00:15:56,840 there's a lot of downstream questions from that. 299 00:15:56,840 --> 00:16:01,120 But like, you know, whether it's a bubble, especially in the AI 300 00:16:01,120 --> 00:16:07,360 infrastructure build out, whether AI is going to cause 301 00:16:07,360 --> 00:16:12,640 bunch of massive job losses and anytime soon, you know, how 302 00:16:12,640 --> 00:16:16,520 does, how does the the music stop or how does, how does this 303 00:16:16,520 --> 00:16:21,520 whole thing wind down? And if you didn't have these 304 00:16:21,520 --> 00:16:26,440 massive data center build outs, would the entire political 305 00:16:26,440 --> 00:16:28,120 situation in the US be different? 306 00:16:28,120 --> 00:16:32,240 Would you know, the right, you know, would we be in a 307 00:16:32,240 --> 00:16:35,240 recession? And then also like to what 308 00:16:35,240 --> 00:16:38,840 degree is there, to what degree is there round tripping of 309 00:16:38,840 --> 00:16:43,480 investments between all these companies that is risky? 310 00:16:44,040 --> 00:16:47,000 I mean, the debt stuff is getting crazy huge amounts of 311 00:16:47,000 --> 00:16:49,520 money. You know, the first Cerebral 312 00:16:49,520 --> 00:16:54,320 Valley was March 2023. And there if you were in that 313 00:16:54,320 --> 00:17:00,040 room, I should have been more entrepreneurial or, but you, you 314 00:17:00,040 --> 00:17:01,520 could have made a lot of money just being like, oh, these 315 00:17:01,520 --> 00:17:03,040 people really see where the world's going. 316 00:17:03,040 --> 00:17:06,680 Let's like invest in NVIDIA. You know, people did make a lot 317 00:17:06,680 --> 00:17:10,880 of money doing that, you know, as you'll sort of see from our 318 00:17:10,880 --> 00:17:13,359 draft picks later in the episode anyway. 319 00:17:14,800 --> 00:17:17,599 But I think this I think this one, I don't think this is 320 00:17:17,640 --> 00:17:20,400 necessarily the end of the bubbleimean.com ran for years 321 00:17:20,400 --> 00:17:22,480 and years. And obviously like, you know, 322 00:17:22,480 --> 00:17:25,240 some of the things sustained in bubbles, you know, it's like you 323 00:17:25,240 --> 00:17:27,720 people start shouting bubble when you're when you're at 324 00:17:27,720 --> 00:17:33,600 evaluation that might not, you know, they might be below where 325 00:17:33,600 --> 00:17:36,600 the bubble ultimately falls too. So it's like there's a run up 326 00:17:36,600 --> 00:17:38,920 and then there's a big fall. People hate the fall. 327 00:17:38,920 --> 00:17:43,000 So I don't know where we are, but I do think this one, it'll 328 00:17:43,000 --> 00:17:47,680 be interesting to see who plays the role of like cheerleader. 329 00:17:47,680 --> 00:17:49,120 It's like, yeah, things are good. 330 00:17:49,120 --> 00:17:51,480 Like don't the party, we need this party to keep going here. 331 00:17:51,680 --> 00:17:54,000 You know, this sort of, there'll be the true believer 332 00:17:54,000 --> 00:17:55,400 cheerleader. There'll be this sort of 333 00:17:55,400 --> 00:17:58,080 disingenuous cheerleader where it's sort of like my job in this 334 00:17:58,080 --> 00:18:01,400 whole business and my employees hang on me saying this thing's 335 00:18:01,400 --> 00:18:03,400 great. And then I think, but I think we 336 00:18:03,400 --> 00:18:06,720 attract and I think this will be a large segment of sort of the 337 00:18:06,720 --> 00:18:08,560 realists. And this is one of the reasons I 338 00:18:08,560 --> 00:18:11,640 like AI more than crypto. I think it's less cult like. 339 00:18:11,640 --> 00:18:14,680 And I think there are lots of people that are like, I'm doing 340 00:18:14,680 --> 00:18:17,880 my smart enterprise thing. We're selling to reasonable 341 00:18:17,880 --> 00:18:20,960 customers. And yes, obviously I'm happy 342 00:18:20,960 --> 00:18:25,600 that open AI might get, you know, 10% better because of 343 00:18:26,000 --> 00:18:28,280 wasteful, you know, infrastructure spending, but 344 00:18:28,280 --> 00:18:30,000 it's not necessarily how I'd run my business. 345 00:18:30,000 --> 00:18:31,960 So I I think there'll be voices of caution. 346 00:18:31,960 --> 00:18:34,240 I don't know what do you guys expect I guess in terms of. 347 00:18:36,120 --> 00:18:38,120 Hype or bubble? Bubble caution. 348 00:18:38,320 --> 00:18:41,000 I mean, I do think it's become like the intellectually en vogue 349 00:18:41,000 --> 00:18:43,440 thing to like talk about the bubble and like complain about 350 00:18:43,440 --> 00:18:47,400 the bubble and everything. I'm not totally sure that 351 00:18:47,400 --> 00:18:53,560 anyone's like investments are really indicating that they 352 00:18:53,560 --> 00:18:56,000 believe there's really a bubble or that they believe it's going 353 00:18:56,000 --> 00:18:57,520 to pop anytime soon, I guess would be. 354 00:18:57,640 --> 00:19:00,520 Investments in terms of venture funds companies venture. 355 00:19:00,520 --> 00:19:05,480 Funds LP's, you know, to your point about crypto, it's not 356 00:19:05,480 --> 00:19:08,720 sort of all about like exiting with the, you know, with the 357 00:19:08,720 --> 00:19:11,560 bags before everyone else like it gets scammed, right? 358 00:19:11,800 --> 00:19:14,600 Like I don't think people are selling out of these really hot 359 00:19:14,600 --> 00:19:19,280 companies very much and I don't think that people are, you know, 360 00:19:19,440 --> 00:19:22,640 not doing the next round at the next craziest evaluation yet in 361 00:19:22,640 --> 00:19:26,680 any meaningful way, right. I think that, I mean, Tyler 362 00:19:26,680 --> 00:19:29,280 Cowen always has this comment about how, well, if you actually 363 00:19:29,280 --> 00:19:32,160 believe the US political system is gonna collapse, why aren't 364 00:19:32,160 --> 00:19:34,480 you short the market? Basically, it's a big comeback, 365 00:19:34,480 --> 00:19:35,880 right? And it's like, there are 366 00:19:35,880 --> 00:19:37,400 different ways to analyze that question. 367 00:19:37,400 --> 00:19:40,000 But I think that in AI, it's like, okay, well, if you 368 00:19:40,000 --> 00:19:43,280 actually believe we're near the peak of a bubble, you can't 369 00:19:43,280 --> 00:19:46,000 really be short necessarily, but you could be secondary ING out. 370 00:19:46,000 --> 00:19:49,040 You could be not participating in the next hot funding round. 371 00:19:49,040 --> 00:19:51,520 You could like, you know, you could be doing various things 372 00:19:51,520 --> 00:19:53,320 that would indicate that you actually believe we're near the 373 00:19:53,320 --> 00:19:54,800 top. And it doesn't really seem like 374 00:19:54,800 --> 00:19:56,360 anyone's doing that. I mean, you're more of an 375 00:19:56,360 --> 00:19:58,960 insider than we are, Eric, but that from the word on the 376 00:19:58,960 --> 00:20:01,600 street, that's not what. I'm my, I think this thing's 377 00:20:01,600 --> 00:20:05,320 still got juice, you know, and I because I talked to, I talked to 378 00:20:05,760 --> 00:20:09,560 skeptical people, you know, I have, you know, lots of the 379 00:20:09,560 --> 00:20:11,400 investors I like to talk to investors who. 380 00:20:11,880 --> 00:20:15,600 Do care about market timing and sort of say, Oh yeah, you needed 381 00:20:15,600 --> 00:20:18,320 to not deploy a ton in 2021 if you want to make a bunch of 382 00:20:18,320 --> 00:20:19,880 money. You know, like those types of 383 00:20:19,880 --> 00:20:24,560 investors are super useful. And like, I think they're like, 384 00:20:24,640 --> 00:20:29,520 the party's still still roaring, you know, I mean, just, you know 385 00:20:29,880 --> 00:20:31,640 what? Amazon's laying off like 10% of 386 00:20:31,640 --> 00:20:36,680 their corporate staff. Like if it does, if new jobs are 387 00:20:36,680 --> 00:20:40,480 going to AI, if consultants are being taken by AI, you know, 388 00:20:40,480 --> 00:20:44,920 like there there's you can do huge market sizing on this 389 00:20:44,920 --> 00:20:48,720 stuff. There's a funny case study I saw 390 00:20:49,480 --> 00:20:51,480 this is one of these it was on Twitter things. 391 00:20:51,480 --> 00:20:52,920 So I should be a little more responsible. 392 00:20:52,920 --> 00:20:56,320 But somebody was like some, you know, a company was going to buy 393 00:20:56,800 --> 00:21:00,800 AI startup and then the consultants like as part of the 394 00:21:00,800 --> 00:21:05,800 acquisition made a demo product that was like supposedly as good 395 00:21:06,200 --> 00:21:10,800 as the start up in like 2 weeks. And I is that pro AI or anti 396 00:21:11,520 --> 00:21:13,480 anti AI? Because on the one hand it's 397 00:21:13,480 --> 00:21:16,040 like, OK, you don't need to buy this. 398 00:21:16,240 --> 00:21:18,360 You know, there are lots of like fake companies that could be 399 00:21:18,360 --> 00:21:19,200 bought. On the other hand, it's like, 400 00:21:19,200 --> 00:21:22,080 oh, you think it's a valuable thing that can be built in just 401 00:21:22,080 --> 00:21:24,080 two weeks? Like it. 402 00:21:24,240 --> 00:21:25,720 Yeah, I don't know. It seems like there's a lot of 403 00:21:25,720 --> 00:21:28,480 stuff stuff coming. Well, yeah, I mean, the numbers 404 00:21:28,480 --> 00:21:31,040 that people are able to raise for compute and scaling and 405 00:21:31,040 --> 00:21:32,280 everything are still just out of control. 406 00:21:32,280 --> 00:21:35,120 I mean, there was that like rumor 18 months ago. 407 00:21:35,120 --> 00:21:36,960 Remember that? Like Sam Alman was trying to 408 00:21:36,960 --> 00:21:40,320 raise $7 trillion for scaling, right? 409 00:21:40,600 --> 00:21:42,040 And people were like, ha, ha, ha. 410 00:21:42,040 --> 00:21:46,160 That's so ridiculous. Like $7 trillion is like, is 411 00:21:46,160 --> 00:21:51,360 like 1/3 of USGDP. It's like 1/3 of the US economic 412 00:21:51,360 --> 00:21:52,760 output. It's just like Sam Alman 413 00:21:52,760 --> 00:21:55,240 personally, it's gonna like fundraise 1/3 of U.S. economic 414 00:21:55,240 --> 00:22:00,040 output for some some microchips. And then I saw someone check in 415 00:22:00,040 --> 00:22:02,120 on this recently and they're like, well, if you put together 416 00:22:02,200 --> 00:22:04,760 all these like Stargate commitments, NVIDIA and AMD and 417 00:22:04,760 --> 00:22:06,880 Broadcom and all this shit, they're like they're up to like 418 00:22:07,080 --> 00:22:09,520 1.5 trillion now. And it's. 419 00:22:09,560 --> 00:22:11,360 Like right? And Sam Ullman, I think he 420 00:22:11,480 --> 00:22:15,000 basically said like, yeah, the 7 trillion people thought that was 421 00:22:15,000 --> 00:22:16,560 too absurd. So I had to like, come up with 422 00:22:16,560 --> 00:22:19,240 like a smaller one that people, you know, it's like clearly it's 423 00:22:19,240 --> 00:22:21,120 just like, yeah, we need a lot of money. 424 00:22:21,560 --> 00:22:25,440 I'm trying to reset expectations. 1.5 trillion in 425 00:22:25,440 --> 00:22:29,240 fundraising right over whatever 12 to 18 months. 426 00:22:29,240 --> 00:22:31,360 Yeah, there are a lot. It's also sloppy. 427 00:22:31,360 --> 00:22:33,240 Like sure of. Course, but still a lot. 428 00:22:33,240 --> 00:22:34,960 I think it's big, Yeah, we can all agree. 429 00:22:34,960 --> 00:22:37,760 That and definitely the economy is, you know, the stock market 430 00:22:37,760 --> 00:22:40,760 is very is in, you know, upon it, yeah. 431 00:22:40,960 --> 00:22:43,760 We've just never seen this level of fundraising in any of our 432 00:22:43,760 --> 00:22:46,000 working lifetimes, right? You know? 433 00:22:46,480 --> 00:22:48,640 So TBD if that means they can only go. 434 00:22:49,000 --> 00:22:50,680 Higher you saw the cursor. But. 435 00:22:51,400 --> 00:22:54,360 This is this is the the ultimate insider bubble work. 436 00:22:54,680 --> 00:22:57,240 The cursor lost one of its Co founders or I do think there are 437 00:22:57,240 --> 00:23:00,880 some of these where it's like such high flying valuations they 438 00:23:00,880 --> 00:23:03,900 raise around and then like they're going to lose a Co 439 00:23:03,900 --> 00:23:05,880 founder. I mean, I don't know the inside 440 00:23:05,880 --> 00:23:09,320 story there, but there, there clearly are going to be, you 441 00:23:09,320 --> 00:23:12,360 know, some of these AI startups that get to raise on. 442 00:23:13,560 --> 00:23:15,840 It's a good, good time to be a journalist, though this doesn't 443 00:23:15,840 --> 00:23:18,160 happen A lot of juicy stories that we'll. 444 00:23:18,280 --> 00:23:23,880 Log on, yeah. I like that your sign of the the 445 00:23:23,880 --> 00:23:27,360 the party continuing, Eric, is the more layoffs we we see in 446 00:23:27,360 --> 00:23:30,760 the US economy. They're pretty hilarious. 447 00:23:30,760 --> 00:23:32,920 They're driven by AI, right? I don't know. 448 00:23:32,960 --> 00:23:35,960 That's a good question. Like if it's Amazon, you know? 449 00:23:37,200 --> 00:23:39,960 Yeah, or or is it just? I do have like, I mean, I do 450 00:23:39,960 --> 00:23:43,080 have two more like sort of real themes I actually think did 451 00:23:43,080 --> 00:23:44,440 happen this year, if that's helpful. 452 00:23:44,640 --> 00:23:47,360 And try to stick with the stick with the term. 453 00:23:47,360 --> 00:23:50,160 So I called mine Fiction Frenzy, I think. 454 00:23:50,280 --> 00:23:55,240 I told mine I called mine AI. Is AI holding up the US economy? 455 00:23:56,600 --> 00:23:57,080 Too. Long. 456 00:23:57,400 --> 00:23:58,040 Too long. Well. 457 00:23:58,040 --> 00:24:02,320 Mine are. Yeah, U.S. economy hangs in the 458 00:24:02,320 --> 00:24:03,720 balance. Yeah, anyway. 459 00:24:03,840 --> 00:24:06,680 My themes are all internally consistent, which is they're all 460 00:24:06,680 --> 00:24:09,600 the year of X, but it's is it the year of X? 461 00:24:09,600 --> 00:24:12,040 We really didn't coordinate on editorial style. 462 00:24:12,040 --> 00:24:15,480 All right, Max. So, so I had, I had the Year of 463 00:24:15,480 --> 00:24:18,080 Agents and the Year of Voice both as being kind of bullshit 464 00:24:18,080 --> 00:24:20,600 themes, but I think there are fake themes. 465 00:24:20,600 --> 00:24:23,640 Yeah, there are two themes that I think were very real themes 466 00:24:23,640 --> 00:24:27,640 this year and saw, you know, real contributions to the way 467 00:24:27,640 --> 00:24:29,800 people work. Maybe ultimately to Amazon 468 00:24:29,800 --> 00:24:33,800 layoffs, like, I don't know. But I think, you know, the the 469 00:24:33,800 --> 00:24:37,400 number one thing to me that was sort of thematically important 470 00:24:37,400 --> 00:24:40,440 this year was the year of the thinking models, right, the 471 00:24:40,440 --> 00:24:43,400 reasoning models. And I know this isn't new news 472 00:24:43,400 --> 00:24:47,280 anymore, but you know, O one was only released like 11-12 months 473 00:24:47,280 --> 00:24:50,160 ago O3, which is the first thinking model that could 474 00:24:50,160 --> 00:24:52,840 incorporate links and actually start referencing real 475 00:24:52,840 --> 00:24:54,880 information. And then ultimately, you know, 476 00:24:54,880 --> 00:24:58,280 we got GPT 5 with the thinking version of that, you know, 477 00:24:58,280 --> 00:24:59,840 Claude Opus, all this kind of stuff, right? 478 00:25:00,040 --> 00:25:02,720 We, we actually made this dramatic architectural 479 00:25:02,720 --> 00:25:06,320 transition in, in terms of how LLMS work in the last, you know, 480 00:25:06,320 --> 00:25:10,240 6 to 12 months that I think took them to a whole nother level of 481 00:25:10,240 --> 00:25:13,040 intelligence and utility. And at least in my personal 482 00:25:13,040 --> 00:25:15,720 life, you know, as a startup founder, whether it's, you know, 483 00:25:15,720 --> 00:25:19,840 doing product planning or writing memos or researching 484 00:25:19,840 --> 00:25:22,520 things or writing emails or whatever, you know, any of the 485 00:25:22,520 --> 00:25:24,280 million things, you know, you could use an LM for. 486 00:25:24,280 --> 00:25:27,720 I just, I feel like my personal usage has gone up like 10X since 487 00:25:27,920 --> 00:25:30,880 3 launched essentially in March or April because I feel like the 488 00:25:30,880 --> 00:25:34,960 utility of these models with thinking combined with sourcing 489 00:25:36,240 --> 00:25:40,600 leads me to find them. I've seen an inflection point in 490 00:25:40,600 --> 00:25:42,680 the value I'm getting out of these models. 491 00:25:43,040 --> 00:25:45,920 And I think that, you know, if anything sort of real happened 492 00:25:45,920 --> 00:25:48,880 to this year, I, I think it was the transition to thinking 493 00:25:48,880 --> 00:25:52,280 models that could, you know, reference, make references and, 494 00:25:52,280 --> 00:25:54,480 and actually sort of be accurate in a lot of ways since. 495 00:25:54,720 --> 00:25:56,960 And maybe that's what's leading to these job losses or these 496 00:25:57,200 --> 00:25:58,800 crazy things you could do with these models now. 497 00:25:59,280 --> 00:26:04,560 Astute No, no argument. I mean it it it's just how hey, 498 00:26:04,640 --> 00:26:06,960 I move so fast. I'm like old news and you know, 499 00:26:06,960 --> 00:26:08,920 like what, 11 months ago? And it's true. 500 00:26:09,000 --> 00:26:11,880 It's. Transformational in a 12 month 501 00:26:11,880 --> 00:26:14,680 time frame, so. Transformation London. 502 00:26:14,760 --> 00:26:17,240 I mean again, I I consider O3 just. 503 00:26:17,560 --> 00:26:20,240 Even O3 which is less like 8 I think 8 months or. 504 00:26:20,240 --> 00:26:22,200 Something. Yeah, yeah, eight months ago, 505 00:26:22,200 --> 00:26:24,880 right, So. And this was one of Carpathy's 506 00:26:25,840 --> 00:26:28,600 sub points, right? That there, there have been 507 00:26:28,600 --> 00:26:31,080 these improvements and we still need to link all this data and 508 00:26:31,080 --> 00:26:34,000 get all this. I mean, it's the, the theme 509 00:26:34,000 --> 00:26:37,280 every year is sort of like, man, there, there's a lot of value 510 00:26:37,280 --> 00:26:41,080 right there that we have that feels like we're not explaining, 511 00:26:41,120 --> 00:26:42,840 you know, here's how you should use it. 512 00:26:43,160 --> 00:26:46,480 I mean, at Dave's sex Medicina, the healthcare health event, we 513 00:26:46,480 --> 00:26:50,120 did, you know, some of the some of these companies, I don't 514 00:26:50,120 --> 00:26:52,920 know, are they that different than Chachi BT but they're 515 00:26:52,920 --> 00:26:56,120 wrapping it in like doctors feel comfortable with it. 516 00:26:56,120 --> 00:26:58,640 They're showing doctors the sources, they're making it work 517 00:26:58,640 --> 00:27:01,200 for medicine. And like, this wasn't one of my 518 00:27:01,200 --> 00:27:04,720 themes, but medicine is one of the areas where AI is going 519 00:27:04,720 --> 00:27:06,840 super hot. Yeah. 520 00:27:06,840 --> 00:27:07,360 And. And so. 521 00:27:07,400 --> 00:27:09,480 Yeah. I think, yeah, I think, I think 522 00:27:09,480 --> 00:27:13,120 similarly, you know, Harvey as the legal AI, the sort of, you 523 00:27:13,120 --> 00:27:16,760 know, scuttlebutt 2 years ago was that it, it actually sucks 524 00:27:16,760 --> 00:27:19,320 and it's not very good, you know, as a legal model. 525 00:27:19,320 --> 00:27:22,880 And then I do think basically it got a brain transplant from 526 00:27:22,880 --> 00:27:25,200 these thinking models in the last 9 to 12 months. 527 00:27:25,200 --> 00:27:27,000 And I mean, my parents are both lawyers. 528 00:27:27,000 --> 00:27:29,120 My dad's like, this is awesome. Like, I mean, like, it's 529 00:27:29,120 --> 00:27:31,240 amazing. You know, I just feel like the 530 00:27:31,280 --> 00:27:34,440 the buzz you're hearing on Harvey or similar legal models 531 00:27:34,440 --> 00:27:36,600 is just way more positive in the last 6 to 12 months. 532 00:27:37,000 --> 00:27:40,400 And I would argue it's all kind of downstream of how good these 533 00:27:40,400 --> 00:27:43,960 thinking models are. I mean, I think ultimately it 534 00:27:43,960 --> 00:27:48,800 gets into this question of if you have a text in, text out 535 00:27:48,800 --> 00:27:54,240 only model, right? You know, do we reach AGI in two 536 00:27:54,240 --> 00:27:56,240 to three years, right? Because I think that I think 537 00:27:56,240 --> 00:28:01,320 that I'm pretty confident that we're nowhere near AGI when it 538 00:28:01,320 --> 00:28:06,040 comes to visual interpretation, when it comes to multi step 539 00:28:06,040 --> 00:28:10,120 processes, when it comes to audio and video, I think we 540 00:28:10,120 --> 00:28:12,640 haven't really reached sort of quote UN quote AGI levels, 541 00:28:12,680 --> 00:28:14,560 right? I mean, James and I were talking 542 00:28:14,560 --> 00:28:17,480 about recently like, you know, I do these little basic visual 543 00:28:17,480 --> 00:28:20,280 tests with GBT thinking, you know, the most advanced model. 544 00:28:20,520 --> 00:28:23,680 I took a picture of a bunch of pomegranate seeds on a plate 545 00:28:23,680 --> 00:28:25,840 because my daughter's willing to pomegranate seeds right now, 546 00:28:26,120 --> 00:28:27,880 because we counted all the pomegranate seeds and there were 547 00:28:27,880 --> 00:28:31,040 like almost exactly 100 pomegranate seeds on this plate. 548 00:28:31,040 --> 00:28:33,280 You know, it was like 99. And I took a picture and it was 549 00:28:33,280 --> 00:28:36,080 like, there are 76 pomegranate seeds on the plate. 550 00:28:36,080 --> 00:28:37,640 And I'm like, no, they're not like we counted. 551 00:28:37,640 --> 00:28:39,480 There's like, you know, a hundred, 101. 552 00:28:39,640 --> 00:28:41,280 And then I was like, hey, we counted. 553 00:28:41,280 --> 00:28:43,120 There's like 101. It's like, you're right. 554 00:28:43,240 --> 00:28:45,680 It's probably a little more. I think it's like 82 pomegranate 555 00:28:45,680 --> 00:28:48,400 seeds, you know? And you're just like, OK, it's 556 00:28:48,400 --> 00:28:51,320 not 82, it's not 75. You're like like my 3 year old 557 00:28:51,360 --> 00:28:54,080 can count pomegranate seeds better than AGI or whatever you 558 00:28:54,080 --> 00:28:54,520 want to. Call it. 559 00:28:54,520 --> 00:28:57,160 Do you think? Do you think that like will get 560 00:28:57,160 --> 00:29:01,600 solved by models or do you think that will get solved by tool 561 00:29:01,600 --> 00:29:02,040 use? Right? 562 00:29:02,040 --> 00:29:05,960 Where you could imagine that GBT doesn't have a connective tool 563 00:29:05,960 --> 00:29:10,960 to the exact right ML model that can count pomegranate seeds, but 564 00:29:12,080 --> 00:29:15,280 like, it probably exists. Deeply wrong that it's so 565 00:29:15,280 --> 00:29:17,280 confident when it doesn't have the tools. 566 00:29:17,320 --> 00:29:21,080 So totally something is wrong there, like if you were just 567 00:29:21,080 --> 00:29:23,120 dealing with another human or another. 568 00:29:23,160 --> 00:29:25,560 They would say like, I can't count, I can't count. 569 00:29:25,840 --> 00:29:27,360 They would just tell you. You don't know. 570 00:29:27,360 --> 00:29:28,680 You know you get it wrong all the time. 571 00:29:28,680 --> 00:29:30,360 People are shouting at you. They get it wrong. 572 00:29:30,360 --> 00:29:33,480 Like no, it speaks to some breaking something. 573 00:29:33,680 --> 00:29:35,280 Wrong. It was like this same day there 574 00:29:35,280 --> 00:29:37,520 was some tweet about how it got a gold medal on the Math 575 00:29:37,520 --> 00:29:40,000 Olympiad or whatever, which is like the hardest math test we've 576 00:29:40,000 --> 00:29:42,160 invented for human beings. And then I'm like, it can't 577 00:29:42,160 --> 00:29:43,280 count. Fucking pomegranate sea. 578 00:29:43,280 --> 00:29:45,600 It's like, it's like, it's like worse than a three-year old. 579 00:29:45,880 --> 00:29:48,760 This was solved by Justin Hoffman in The Rain Man Like. 580 00:29:49,400 --> 00:29:53,040 This yeah. So my point to come back to this 581 00:29:53,040 --> 00:29:56,760 is that I think for visual audio tool use whatever, you know, 582 00:29:56,760 --> 00:29:59,520 James's category, there's just a long way to go in a lot of these 583 00:29:59,520 --> 00:30:01,520 areas, right? But I think if you're like, OK, 584 00:30:01,520 --> 00:30:04,760 text in, text out, you know, we're already at this gold 585 00:30:04,760 --> 00:30:06,840 medal, you know, US math Olympiad level. 586 00:30:06,840 --> 00:30:11,280 I think that, you know, O5O6O7, you know, I know they renamed 587 00:30:11,280 --> 00:30:13,400 all these fucking things. So whatever numbering scheme you 588 00:30:13,400 --> 00:30:15,360 want to take. But the next two or three 589 00:30:15,360 --> 00:30:20,520 generations of models, I can see us getting to a text in, text 590 00:30:20,520 --> 00:30:23,080 out, you know, AGI level of thinking, right? 591 00:30:23,280 --> 00:30:27,960 Where I for me, the leap forward we took between GBT 4-O and O3 592 00:30:27,960 --> 00:30:32,360 was the first time that I really believed, you know, I got AGI 593 00:30:32,360 --> 00:30:35,440 pill that like, Oh yeah. As at least as a chat bot, this 594 00:30:35,440 --> 00:30:38,520 thing is going to be, you know, much, much smarter than the 595 00:30:38,520 --> 00:30:41,160 average human and basically able to do anything and potentially 596 00:30:41,160 --> 00:30:43,640 break ground in terms of scientific discoveries and other 597 00:30:43,640 --> 00:30:45,680 things like that. Related theme, My last theme 598 00:30:45,880 --> 00:30:47,800 related. And I just wanted to add it on 599 00:30:47,800 --> 00:30:52,480 as a sort of postscript. We sort of forget again another 600 00:30:52,480 --> 00:30:56,040 old news thing that the term vibe coding was coined this 601 00:30:56,040 --> 00:30:58,480 year. The idea? 602 00:30:58,960 --> 00:31:01,200 Yeah, what? Karpati's tweet was like January 603 00:31:01,200 --> 00:31:03,720 this year or something? January is not this year. 604 00:31:03,720 --> 00:31:04,600 That's like a different. Year. 605 00:31:04,600 --> 00:31:08,320 No, it was it. Was this year? 606 00:31:08,320 --> 00:31:11,040 Isn't it crazy? Like, so I would just say as a 607 00:31:11,040 --> 00:31:14,200 postscript, as an extension of the thinking model progress and 608 00:31:14,200 --> 00:31:17,240 as a post, an extension of what we've seen with these text in 609 00:31:17,240 --> 00:31:19,760 textile models. And I would call this the year 610 00:31:19,760 --> 00:31:22,720 of vibe coding, right? I mean, this is the clearly the 611 00:31:22,720 --> 00:31:25,480 mimetic phrase of the year. There's a million startups that 612 00:31:25,480 --> 00:31:26,840 are getting crazy evaluations off of. 613 00:31:26,840 --> 00:31:28,720 It, it, it is. And we were having Guillermo at 614 00:31:28,720 --> 00:31:32,440 Versal and Amjad at Repplit, both on stage, different panels. 615 00:31:32,600 --> 00:31:35,480 We had lovable at CV London, you know, exactly right. 616 00:31:35,760 --> 00:31:38,520 I would say internally to our company that all of our 617 00:31:38,520 --> 00:31:41,640 designers and product managers now believe they can code, you 618 00:31:41,640 --> 00:31:44,960 know, websites and prototype things for internal tools like 619 00:31:45,280 --> 00:31:47,040 we have. Clearly, you know. 620 00:31:47,040 --> 00:31:50,640 Cerebral Valleys. Logos 3D AI generated or like 621 00:31:50,640 --> 00:31:53,880 the design was human, but I think the animation was yeah. 622 00:31:54,120 --> 00:31:56,920 So I would just say vibe coding as a phrase that has entered our 623 00:31:56,920 --> 00:32:00,160 lives, you know, representing the idea of telling the coding 624 00:32:00,160 --> 00:32:02,600 bot to what to do and then getting back results and sort of 625 00:32:03,120 --> 00:32:07,440 riding the vibes, as Karpati said when he coined it was a 626 00:32:07,440 --> 00:32:11,760 big, big deal and I think was a huge part of what we actually 627 00:32:11,760 --> 00:32:15,080 saw happen in the world this year based on large models in 628 00:32:15,120 --> 00:32:16,720 AI. So yeah, those are my 2. 629 00:32:16,840 --> 00:32:21,120 I love how your strategy is look at the dates and be like some of 630 00:32:21,120 --> 00:32:24,880 these things we take for granted are we were so greedy. 631 00:32:24,880 --> 00:32:27,000 You know, it's like a lot has happened and it's still very 632 00:32:27,000 --> 00:32:28,880 current. And so we're gonna be unpacking 633 00:32:29,240 --> 00:32:30,800 those those themes. Well, well. 634 00:32:31,160 --> 00:32:32,800 You spent all this time complaining about things that 635 00:32:32,800 --> 00:32:34,240 were too early. I'm. 636 00:32:34,240 --> 00:32:36,800 Saying no, no, no, no, I like it actually here. 637 00:32:36,800 --> 00:32:39,120 I think it's it's a good point and you're differentiating 638 00:32:39,120 --> 00:32:41,240 obviously agents and other stuff. 639 00:32:41,240 --> 00:32:43,000 You know, we got I'm not I'm not criticizing. 640 00:32:43,000 --> 00:32:44,320 This is just funny. Actually worked. 641 00:32:44,760 --> 00:32:48,440 This year in my opinion, right? Yes, and date them to 2025. 642 00:32:48,640 --> 00:32:53,680 Yeah, James. Yeah, I, I have a, a tangent out 643 00:32:53,680 --> 00:33:00,000 of that theme for Max, I guess I'm calling it the the jagged 644 00:33:00,000 --> 00:33:04,520 edge of AI. Essentially, if you're not 645 00:33:04,520 --> 00:33:08,040 someone who's using these coding models or vibe coding all the 646 00:33:08,040 --> 00:33:14,560 time or playing with the latest VO model like or Sora like, 647 00:33:14,560 --> 00:33:19,600 you're just not really seeing the dramatic improvements. 648 00:33:19,960 --> 00:33:23,800 So I think most people in the world are are, you know, maybe 649 00:33:23,800 --> 00:33:28,160 using chat GPG 5 now and getting a little of this thinking and 650 00:33:28,160 --> 00:33:35,560 reasoning, you know, experience, but not much beyond that. 651 00:33:35,560 --> 00:33:38,960 And then they are thinking, oh, this is going to be another over 652 00:33:38,960 --> 00:33:45,440 promising technology. This, you know, is kind of a 653 00:33:45,440 --> 00:33:50,240 Silicon Valley pipe dream AGI. And then what you hear from the 654 00:33:50,240 --> 00:33:54,480 labs themselves is like, Nope, everything's going pretty well. 655 00:33:54,480 --> 00:33:58,000 Like on track to AGI, you know, we're gonna have automated 656 00:33:58,000 --> 00:34:01,080 science in a couple of years. And but Dario made some 657 00:34:01,080 --> 00:34:03,480 predictions that aren't really holding up right, like some of 658 00:34:03,480 --> 00:34:04,880 these predictions. I think that. 659 00:34:04,880 --> 00:34:08,000 What, didn't you say 90% of code was gonna be written six months 660 00:34:08,040 --> 00:34:10,760 ago? I think that's not that far off 661 00:34:10,760 --> 00:34:12,920 inside of anthropic like I think like. 662 00:34:12,920 --> 00:34:16,159 This is like self driving. Just like definition question or 663 00:34:16,159 --> 00:34:18,120 something or? No, no, I self driving where is 664 00:34:18,120 --> 00:34:19,360 just like tomorrow, tomorrow, tomorrow. 665 00:34:19,360 --> 00:34:21,760 We're almost there. I'm literally saying like I 666 00:34:21,760 --> 00:34:24,639 think today you could would make an argument based on what I've 667 00:34:24,639 --> 00:34:28,840 heard from Anthropic that you know 90% of the code is written. 668 00:34:28,880 --> 00:34:30,040 You're saying it's a human problem. 669 00:34:30,040 --> 00:34:32,480 The humans just haven't really figured out how to use the good 670 00:34:32,480 --> 00:34:34,159 technology. But it's like positive. 671 00:34:34,320 --> 00:34:38,440 Yeah, yeah, exactly. Like and I you hear from open AI 672 00:34:38,440 --> 00:34:43,159 researchers and engineers that they like vibe coded or you 673 00:34:43,159 --> 00:34:46,920 know, used codecs, their own, you know, version of cloud code 674 00:34:47,080 --> 00:34:49,679 to build like a lot of the products that came out at 675 00:34:49,679 --> 00:34:53,120 Devday. So I actually think there's just 676 00:34:53,120 --> 00:34:58,120 like more adoption happening within labs than people realize 677 00:34:58,120 --> 00:35:02,000 and like the people who are AGI pilled in lat in the labs have 678 00:35:02,000 --> 00:35:05,200 like a good reason to be and and we're not seeing all of that. 679 00:35:05,360 --> 00:35:07,400 Yeah. I mean, I, I agree with that. 680 00:35:07,520 --> 00:35:09,400 I agree with the jagged frontier in general. 681 00:35:09,400 --> 00:35:14,160 I think that it's, it's already not evenly distributed how good 682 00:35:14,160 --> 00:35:17,440 these models are at some things. And, and I think particularly I 683 00:35:17,440 --> 00:35:20,720 think the sort of GPD 5 launch to your point was sort of this 684 00:35:21,080 --> 00:35:24,280 anti PR moment where people are like, well, AI like it's not 685 00:35:24,280 --> 00:35:26,200 really any better now. It's just sort of like it's just 686 00:35:26,360 --> 00:35:28,480 kind of sucks. And it's like, well, try 687 00:35:28,480 --> 00:35:30,040 clicking the little. It's insane. 688 00:35:30,120 --> 00:35:34,600 Yeah, yeah, it takes some. You have to do the work, you 689 00:35:34,600 --> 00:35:37,640 know, you have to find the value, right, Chaji? 690 00:35:37,640 --> 00:35:40,320 But it's insane, this picture that you're painting that it's 691 00:35:40,320 --> 00:35:43,360 like, Oh yeah, they just have access to Chaji Buti and they 692 00:35:43,360 --> 00:35:45,000 can't see the value. I'm like what? 693 00:35:45,000 --> 00:35:47,680 Like they're. Do you agree with my my? 694 00:35:47,760 --> 00:35:51,760 Sort of Astoria or version of that, that most people are, 695 00:35:51,800 --> 00:35:55,960 yeah, kind of, yeah, to Max's point, disappointed by the the 696 00:35:55,960 --> 00:35:59,440 ChatGPT 5 launch. Yeah, I think there are a lot of 697 00:35:59,560 --> 00:36:01,760 sort of skeptics. It takes too much work. 698 00:36:01,760 --> 00:36:04,120 Yeah, to get the value out of it. 699 00:36:04,120 --> 00:36:06,400 And people need to be led to water. 700 00:36:06,400 --> 00:36:08,240 That's what that's what I keep saying. 701 00:36:08,240 --> 00:36:09,960 You know, it sort of needs to be handed to them. 702 00:36:09,960 --> 00:36:13,440 Here's the thing it can do. I hear a lot from in even people 703 00:36:13,440 --> 00:36:18,680 in Silicon Valley who say, oh, the labs are lying about AGI. 704 00:36:18,960 --> 00:36:23,640 You know, if you look at what's opening eyes releasing it's, you 705 00:36:23,640 --> 00:36:28,480 know, Sora and they're gonna just hack attention. 706 00:36:28,480 --> 00:36:34,560 And now Sam is talking about erotica in ChatGPT. 707 00:36:34,800 --> 00:36:39,880 So like, you know, it's all, it's all, yeah, it's all we're. 708 00:36:39,880 --> 00:36:43,200 Talking about AGI less right the public commerce, it doesn't feel 709 00:36:43,200 --> 00:36:44,720 like the labs are talking about AGI. 710 00:36:44,760 --> 00:36:46,640 Don't agree with that. I think they, I think they still 711 00:36:46,640 --> 00:36:48,240 talk about it a lot. Like I mean. 712 00:36:48,560 --> 00:36:50,640 I think yeah, I think, I think they we. 713 00:36:50,640 --> 00:36:51,920 Should gauge that I'm. Talking about it. 714 00:36:51,920 --> 00:36:52,720 What? Live streaming? 715 00:36:52,720 --> 00:36:55,280 Yeah, yeah. I mean one thing. 716 00:36:55,320 --> 00:36:57,720 I mean Microsoft right? Like the they just announced 717 00:36:57,720 --> 00:37:00,760 this? Yes, it's contractual. 718 00:37:00,760 --> 00:37:02,200 Yeah. Microsoft had this deal with 719 00:37:02,200 --> 00:37:05,480 Open AI and you know, some of the triggers are some of the 720 00:37:05,480 --> 00:37:08,440 deal terms are conditional on AGI, but in some ways it goes 721 00:37:08,440 --> 00:37:11,920 both ways because it's a deal. So Microsoft could be like AGI 722 00:37:11,920 --> 00:37:13,360 is not close, that's good for us. 723 00:37:13,360 --> 00:37:16,480 We have a better deal, you know, so maybe Open AI thinks it's 724 00:37:16,480 --> 00:37:19,840 close. Now I guess Open AI can declare 725 00:37:19,840 --> 00:37:22,360 AGI to get out of the contract and then there's like an 726 00:37:22,360 --> 00:37:28,520 independent panel of reviewers that will decide if if AGI has 727 00:37:28,520 --> 00:37:30,560 actually been achieved, which I thought was pretty interesting. 728 00:37:30,640 --> 00:37:32,560 Here's the last theme that I'll be quick about. 729 00:37:32,560 --> 00:37:36,960 Don't say EA, like don't say effective altruist. 730 00:37:37,200 --> 00:37:40,080 I mean that not like we're bringing that up all the time, 731 00:37:40,080 --> 00:37:43,400 but I think, you know, anthropic is obviously run away from that 732 00:37:43,400 --> 00:37:45,560 term. You know, we had what we had 733 00:37:45,560 --> 00:37:51,200 Holden right, who is a former kind of yeah, speak at one of 734 00:37:51,200 --> 00:37:52,360 them. So it's been come up. 735 00:37:52,360 --> 00:37:54,000 He's married to Danielle at Anthropic. 736 00:37:54,000 --> 00:37:57,440 She spoke like Anthropic doesn't want to talk about effective 737 00:37:57,480 --> 00:38:00,480 altruism anymore. And then there's this sort of 738 00:38:00,480 --> 00:38:02,960 Trump world sort of thing. You know, you saw David Sachs 739 00:38:02,960 --> 00:38:06,720 fighting with Jack Clark over having any sort of safety ISM. 740 00:38:07,040 --> 00:38:10,600 And so it's beyond just don't say yeah, it's will people talk 741 00:38:10,600 --> 00:38:16,680 about safety at all and how much in sort of a Trump world is even 742 00:38:16,680 --> 00:38:19,960 just worrying about AI safety Lib coded and therefore nobody 743 00:38:19,960 --> 00:38:23,280 wants to talk about it. Yeah, David Sachs went like 744 00:38:23,280 --> 00:38:29,080 further and said the all of this was regulatory capture, which I 745 00:38:29,080 --> 00:38:31,440 thought was just a very. There was like performative. 746 00:38:31,440 --> 00:38:34,360 Jump like, like it's clearly not performative. 747 00:38:34,360 --> 00:38:36,080 To your point. Like they've been talking about 748 00:38:36,080 --> 00:38:39,000 this for a decade. Everything he does is 749 00:38:39,000 --> 00:38:41,440 disingenuous. All his arguments are motivated 750 00:38:41,440 --> 00:38:44,720 for some end, so he cannot comprehend someone having. 751 00:38:44,760 --> 00:38:47,080 Yeah, yeah, argument just on principle. 752 00:38:48,200 --> 00:38:50,880 All right, those are our themes. I'm giving myself the last word 753 00:38:50,880 --> 00:38:53,160 on that one. Very excited for this next 754 00:38:53,160 --> 00:38:54,520 segment. We're going to go into the 755 00:38:54,520 --> 00:38:56,680 draft. We've been doing this for a 756 00:38:56,680 --> 00:38:59,960 couple years now, so we already have our picks and we're going 757 00:38:59,960 --> 00:39:03,440 to be sort of dropping and then picking and picking up some new 758 00:39:03,440 --> 00:39:06,440 ones. And I'm very excited to say with 759 00:39:06,440 --> 00:39:09,320 the upped production value on the Newcomer podcast, we have a 760 00:39:09,320 --> 00:39:12,800 hype video to catch you up on what we've done to try and make 761 00:39:12,800 --> 00:39:15,080 this comprehensible. So you can follow along. 762 00:39:15,320 --> 00:39:19,960 This ultimate feat of nerddom, which is taking what people do 763 00:39:19,960 --> 00:39:23,640 for cool things like sports and using it to pick not just AI 764 00:39:23,640 --> 00:39:27,000 nerdy things, but companies. I feel like it's a double whammy 765 00:39:27,000 --> 00:39:31,160 of Loserdum and that we're we're rooting for companies to succeed 766 00:39:31,160 --> 00:39:34,480 and we're interested in in tech. So anyway, here's that hype 767 00:39:34,480 --> 00:39:37,640 video. In living rooms everywhere, die 768 00:39:37,640 --> 00:39:40,600 hard fans chase glory through fantasy football. 769 00:39:40,840 --> 00:39:44,200 In newsrooms, armchair strategist bet the house on who 770 00:39:44,200 --> 00:39:48,720 wins the Oval Office. But in Silicon Valley, these 771 00:39:48,720 --> 00:39:51,840 insiders draft generative AI startups. 772 00:39:52,960 --> 00:39:57,480 This is the AI Fantasy Draft League IN2023G Miss Eric 773 00:39:57,480 --> 00:40:00,880 Newcomer, Max Child and James Wilsterman chose their five 774 00:40:00,880 --> 00:40:03,360 startups. Open AI, Inflection, Character 775 00:40:03,400 --> 00:40:07,480 AI, Glean and Mistral AI. Databricks, Pine Cone, Cohere, 776 00:40:07,480 --> 00:40:11,920 Modular, and Imbue. Hugging Face, Anthropic AI 21 777 00:40:11,920 --> 00:40:15,360 Labs, Replit and Adept. After a year of celebration and 778 00:40:15,360 --> 00:40:16,840 regret. Dominating right now. 779 00:40:16,880 --> 00:40:19,200 Bad pick, bad pick. In like 6 months that has just 780 00:40:19,320 --> 00:40:21,840 totally changed the. GMs went back to the table to 781 00:40:21,840 --> 00:40:25,200 evaluate their teams and prepare to add 2 second round draft 782 00:40:25,200 --> 00:40:29,680 picks with Eric in the lead in 2024, followed by James and Max 783 00:40:29,680 --> 00:40:32,040 struggling to hold on to his franchise death. 784 00:40:32,040 --> 00:40:35,960 Defying public IPOs, market fluctuations and many rounds of 785 00:40:35,960 --> 00:40:39,000 funding later, the boys are back in the hot seat once again. 786 00:40:39,080 --> 00:40:42,320 All in the hopes of glory on November 1st, 2028 when the 787 00:40:42,320 --> 00:40:45,360 Champions will be crowned by total market cap valuation. 788 00:40:45,440 --> 00:40:49,640 Welcome to the 2025 AI Fantasy Draft League, presented in 789 00:40:49,640 --> 00:40:54,640 partnership by Newcomer. Wow, production value is just 790 00:40:55,800 --> 00:40:59,000 just skyrocketed. If if if that had come out 791 00:40:59,000 --> 00:41:01,960 before TBPN existed, I would say we should start. 792 00:41:01,960 --> 00:41:07,880 TBPN, not like TBPN invented production like Youtubers have 793 00:41:07,880 --> 00:41:09,640 been doing this forever. Everybody's yeah. 794 00:41:09,640 --> 00:41:12,960 But anyway, I'm we're hooked. Up ESPN before they did OK. 795 00:41:13,520 --> 00:41:17,960 I know and Riley, my business guy, literally one of his, the 796 00:41:18,040 --> 00:41:21,120 his main editorial insight the whole time he's been here, he's 797 00:41:21,120 --> 00:41:23,080 been with me three years. He's been like, you need to be 798 00:41:23,080 --> 00:41:25,160 more like sports. You need to be more like sports. 799 00:41:25,160 --> 00:41:27,520 He's like and he's like be more correct about. 800 00:41:27,680 --> 00:41:32,160 Yeah, I told. You I take this seriously. 801 00:41:32,160 --> 00:41:34,960 So I feel like the sports thing is a little glib sometimes. 802 00:41:34,960 --> 00:41:37,080 Anyway, I love I love that. That's great. 803 00:41:37,840 --> 00:41:38,880 We should get into it. James. 804 00:41:38,880 --> 00:41:43,560 You're the both a player, but also sort of the game master of 805 00:41:43,600 --> 00:41:45,160 this. I don't know what the. 806 00:41:45,160 --> 00:41:46,440 What do you call it when you run? 807 00:41:46,440 --> 00:41:48,320 Missioner. Isn't there a term commissioner? 808 00:41:48,320 --> 00:41:49,280 Commissioner. Yeah, OK. 809 00:41:49,520 --> 00:41:55,920 Yeah. Yep, this is year three-year 810 00:41:55,920 --> 00:42:02,280 three of the draft. We've been working on this for a 811 00:42:02,280 --> 00:42:05,240 while now. We all have teams just to set 812 00:42:05,240 --> 00:42:08,960 the the rules again. You know, we drafted a few teams 813 00:42:09,040 --> 00:42:11,560 originally. Now there's like waiver pickups 814 00:42:11,560 --> 00:42:13,640 and you can, you can add and drop teams. 815 00:42:14,000 --> 00:42:18,080 We're drafting companies that have raised over $100 million 816 00:42:18,080 --> 00:42:21,600 have a core use of generative AI or a core role in the generative 817 00:42:21,640 --> 00:42:24,880 AI ecosystem. There are some excluded 818 00:42:24,880 --> 00:42:28,560 companies if they're really focused on bio health, defense, 819 00:42:30,120 --> 00:42:33,480 cloud computing or or chips or silicon we. 820 00:42:33,520 --> 00:42:35,920 Block the whole chip infrastructure that which is 821 00:42:35,920 --> 00:42:38,800 sort of funny in retrospect and now that everything's flowing 822 00:42:38,800 --> 00:42:41,280 there, but but we we excluded that. 823 00:42:41,280 --> 00:42:44,520 Yeah, we could always bring it back if you guys are don't think 824 00:42:44,520 --> 00:42:45,840 we have enough companies to expect. 825 00:42:46,640 --> 00:42:49,800 Yeah, I know we're not. We're not focused on robotics 826 00:42:49,800 --> 00:42:53,440 companies or China companies based in China. 827 00:42:54,640 --> 00:42:59,080 And yeah, it's a snake draft this this season, this year. 828 00:42:59,120 --> 00:43:04,280 I always like to say I had to. I insisted that we have an 829 00:43:04,280 --> 00:43:08,320 auction for who goes first, and we auctioned and I paid $75 830 00:43:08,320 --> 00:43:11,800 billion so that I could pick Open AI first because my view 831 00:43:11,800 --> 00:43:14,680 from the beginning was that Open AI was going to swamp it all and 832 00:43:14,680 --> 00:43:16,520 you needed the winner. It's a home run business. 833 00:43:16,680 --> 00:43:19,640 So far that strategy is doing well though, so. 834 00:43:19,640 --> 00:43:23,400 Far, so far, Open AI is worth more than James and my team 835 00:43:23,400 --> 00:43:27,360 combined this year, which is a perfect expression of power law 836 00:43:27,360 --> 00:43:29,720 dynamics. Now you did pay a $75 billion 837 00:43:29,720 --> 00:43:32,280 handicap, so it's a little bit under, but it's pretty damn 838 00:43:32,280 --> 00:43:35,920 close, which is kind of mind. Boggling but it wasn't like oh I 839 00:43:35,920 --> 00:43:38,560 just happened to go first by some random chance I paid for 840 00:43:38,560 --> 00:43:41,880 the right knowing that we would fight for opening and. 841 00:43:41,960 --> 00:43:47,360 You, your team's worth close to 510 billion right now, including 842 00:43:47,360 --> 00:43:52,400 that handicap of 75 billion Max, your team's worth 150 billion 843 00:43:53,360 --> 00:43:56,120 and my team is worth close to 400 billion. 844 00:43:57,520 --> 00:44:01,640 Just to catch you up on last season, we we, we dropped a few 845 00:44:01,640 --> 00:44:05,440 companies and we had some exit. So that opened up new slots on 846 00:44:05,440 --> 00:44:08,440 our teams. I drafted Xai with the first 847 00:44:08,440 --> 00:44:11,520 pick last year and that's been a huge win. 848 00:44:11,680 --> 00:44:14,840 Now rumored to be valued around 200 billion. 849 00:44:15,080 --> 00:44:18,360 Great fall in your lap sort of opportunity, I mean your first. 850 00:44:18,640 --> 00:44:21,360 Is this the moment where I get to say I basically got Coin 851 00:44:21,360 --> 00:44:24,520 flipped out of being in second place because if I had drafted 852 00:44:24,680 --> 00:44:29,240 XAI I would be in second place by a mile just like James for 853 00:44:29,240 --> 00:44:29,640 the. Record. 854 00:44:29,640 --> 00:44:32,720 Fair, fair. Well, fortunately for you Max, 855 00:44:32,720 --> 00:44:35,240 you get to draft first this year. 856 00:44:35,240 --> 00:44:36,680 Yeah, I know. And this is the year where 857 00:44:36,680 --> 00:44:38,720 there's no. Obvious wins. 858 00:44:39,800 --> 00:44:41,640 Yeah, I know. This is oh wow, I get the first 859 00:44:41,640 --> 00:44:43,800 pick of the year where everything's worth $10 billion 860 00:44:43,800 --> 00:44:48,280 and nothing's worth 20. So I guess first to start, Max, 861 00:44:48,840 --> 00:44:52,120 do you want to just quickly go, you know, run through your team 862 00:44:52,120 --> 00:44:55,760 and then also let us know if you're dropping any of your 863 00:44:55,760 --> 00:44:57,080 team? Sure. 864 00:44:57,080 --> 00:45:03,560 My team today is Data Bricks, Cohere Modular Scale, Sierra 865 00:45:03,560 --> 00:45:07,800 Sakana and Hebia. You know, it's not great. 866 00:45:07,800 --> 00:45:12,240 Other than data bricks, I did get an exit with scale at $29 867 00:45:12,240 --> 00:45:15,240 billion. Are we calling that an exit? 868 00:45:15,240 --> 00:45:17,760 We basically I think. I'm calling that we, we've, 869 00:45:17,960 --> 00:45:20,640 we've gone through this a couple times where there's been these, 870 00:45:20,640 --> 00:45:24,160 yeah, pseudo exits count. So I'm counting that. 871 00:45:25,880 --> 00:45:29,560 Yeah, again, I'm in last place by a huge margin. 872 00:45:31,120 --> 00:45:33,720 Not not, you know, my team out on the field has just not really 873 00:45:33,720 --> 00:45:36,760 been performing. Sierra is a good company, but 874 00:45:36,880 --> 00:45:39,760 and Co here I think is gonna have a second I think they're. 875 00:45:39,800 --> 00:45:41,720 Yeah, yeah, I think we are. There's some value here. 876 00:45:41,720 --> 00:45:45,200 I think ultimately, you know, as I said earlier, had I been, had 877 00:45:45,200 --> 00:45:48,720 I won the coin flip and gotten to pick Xai, I would be 878 00:45:48,920 --> 00:45:51,040 breathing down Eric's neck kind of like James. 879 00:45:51,040 --> 00:45:54,200 And instead I'm I'm a joke, you know, looming far in the 880 00:45:54,200 --> 00:45:57,720 background. So with that, I'm gonna do some 881 00:45:57,720 --> 00:46:02,160 aggressive restructuring on my team this year and do some heavy 882 00:46:02,160 --> 00:46:04,760 dropping and try to pick up a lot of value in a hurry. 883 00:46:05,600 --> 00:46:08,680 So I'm out on scale. Yeah, I'm also going to drop 884 00:46:09,360 --> 00:46:12,360 modular. I'm also gonna drop Sakana and 885 00:46:12,360 --> 00:46:16,280 I'm also gonna drop heavier, so I'm gonna make 4 picks in 886 00:46:16,280 --> 00:46:19,480 addition to my 22 picks that I get as part of the draft this 887 00:46:19,480 --> 00:46:21,200 year. So you get 4 picks or? 888 00:46:21,200 --> 00:46:23,880 Yeah, you get 4 + 2. So you get 6 picks? 889 00:46:24,120 --> 00:46:27,720 Yes, that's a lot, yeah. Yes, I'm taking, I'm shooting my 890 00:46:27,720 --> 00:46:31,040 shot here at restructuring. Look, sometimes you gotta have. 891 00:46:31,040 --> 00:46:32,600 A rebuilding year? Yeah, I think. 892 00:46:32,600 --> 00:46:35,160 You gotta you gotta go all in, which is what I'm doing this 893 00:46:35,160 --> 00:46:37,760 year, so we're really going for it. 894 00:46:38,120 --> 00:46:40,920 Eric, you're gonna pick second this year. 895 00:46:41,920 --> 00:46:45,320 Yeah, so my team, first of all, I've dropped nobody. 896 00:46:45,560 --> 00:46:48,760 All my exits are from actual deals. 897 00:46:48,760 --> 00:46:50,560 So I think that's another point in my favor. 898 00:46:50,560 --> 00:46:57,080 But OK, I currently have Open AI, Glean, Mistral, Perplexity, 899 00:46:57,400 --> 00:47:03,520 Safe, Super Intelligence and Harvey Kodium just exited. 900 00:47:03,560 --> 00:47:06,280 So I'm slash windsurf, whatever you want to call them. 901 00:47:07,200 --> 00:47:10,000 So James has that at 2.4 billion. 902 00:47:10,000 --> 00:47:15,320 I've been scrutinized those small potatoes and then I've 903 00:47:15,320 --> 00:47:19,400 already exited inflection and character. 904 00:47:19,560 --> 00:47:21,240 So we I replaced those last time. 905 00:47:21,240 --> 00:47:24,760 So the I have an extra spot thanks to windsurf. 906 00:47:25,600 --> 00:47:28,440 But otherwise, yeah, I'm I'm holding on. 907 00:47:28,440 --> 00:47:31,640 I I like I like mine. I mean, who knows what's 908 00:47:31,640 --> 00:47:33,000 happening at safe super intelligence. 909 00:47:33,000 --> 00:47:36,600 We're just betting on Ilia there perplexity, obviously lots of 910 00:47:37,120 --> 00:47:39,440 zeitgeist glean. You know, I really believe in 911 00:47:39,440 --> 00:47:42,200 the business. I just don't know, you know, I I 912 00:47:42,200 --> 00:47:44,080 think that one has the lowest to fall. 913 00:47:44,080 --> 00:47:46,800 It's just like, does it have the true insanity? 914 00:47:46,800 --> 00:47:49,320 You know, it seems like this this thing is supposed to end 915 00:47:49,320 --> 00:47:53,000 what we said 2028, November. It's like the bubble bubble 916 00:47:53,000 --> 00:47:54,520 could still be going for all we know. 917 00:47:54,520 --> 00:47:58,320 But glean, I feel like bubble or not, I'm, I'm, I'm betting on 918 00:47:58,320 --> 00:48:01,040 that one. Yeah. 919 00:48:01,040 --> 00:48:04,640 So I I like this team in Harvey. You know, we're putting them on 920 00:48:04,640 --> 00:48:05,600 stage at St. Row Valley. 921 00:48:05,600 --> 00:48:07,840 They're getting, they're getting just another boot, the little 922 00:48:07,840 --> 00:48:10,920 boost that they need, you know, to really go to the point. 923 00:48:11,080 --> 00:48:14,640 To your point, it'll be it'll be hilarious if in 2028 where the 924 00:48:14,640 --> 00:48:17,640 bubble has popped and the stock market has collapsed and the 925 00:48:17,640 --> 00:48:20,640 final scoring is really just about like whose businesses 926 00:48:20,640 --> 00:48:24,520 collapsed the least in the. Yeah, bubbles look like human 927 00:48:24,520 --> 00:48:26,840 made this fucking fortune zone at the right time, you know, 928 00:48:26,840 --> 00:48:29,080 like yeah, yeah, your scale deal could be. 929 00:48:29,120 --> 00:48:32,600 Like open AI will be like -400. Billion and like. 930 00:48:32,920 --> 00:48:36,160 You know, James and I all have these like, little like cobbled 931 00:48:36,160 --> 00:48:37,520 together. Stones like. 932 00:48:37,520 --> 00:48:41,040 Companies can only go to 0. You can't be -400 billion. 933 00:48:41,040 --> 00:48:43,320 I don't know what you're. I mean on your current valuation 934 00:48:43,320 --> 00:48:45,520 like from 500 to 1. 100 Oh, I see. 935 00:48:45,560 --> 00:48:46,600 Yeah. I'll draw from you. 936 00:48:46,600 --> 00:48:48,600 I thought, I thought like, I just meant I thought like they 937 00:48:48,600 --> 00:48:52,440 were gonna be in debt, you know, to, to the Saudis or something. 938 00:48:52,480 --> 00:48:53,600 New to that? No. 939 00:48:53,600 --> 00:48:57,400 No, no, I just meant, I just meant 400 down from today, not 940 00:48:57,400 --> 00:48:59,760 not -400 valuation. We'll see. 941 00:49:00,160 --> 00:49:03,960 Yeah, yeah. OK, so 3 picks for Eric, 2 new 942 00:49:03,960 --> 00:49:08,880 picks and 1 exit. My team is Hugging Face, 943 00:49:08,880 --> 00:49:14,680 Anthropic, Replit, XAI Runway 11 labs and poolside. 944 00:49:15,560 --> 00:49:21,800 Getting some value from XAI obviously and Anthropic the rest 945 00:49:22,800 --> 00:49:26,440 sort of small potatoes, but poolside is kind of rumoured at 946 00:49:26,440 --> 00:49:29,480 that $12 billion evaluation recently, which I think is 947 00:49:29,520 --> 00:49:37,680 interesting. 11 Labs just had a tender at 6.6 billion and runway 948 00:49:37,680 --> 00:49:40,720 rumoured at around 5 billion. So some value here. 949 00:49:41,080 --> 00:49:45,120 I'm definitely thinking about dropping Hugging Face this year. 950 00:49:45,120 --> 00:49:47,520 That's I don't have any exits this year so I'm drop. 951 00:49:47,560 --> 00:49:50,320 I'm going to drop Hugging Face. They just haven't raised. 952 00:49:50,320 --> 00:49:54,360 Brutal. They just haven't raised, you 953 00:49:54,360 --> 00:49:56,080 know? I never thought that pick made 954 00:49:56,080 --> 00:49:57,080 sense. We talked. 955 00:49:57,080 --> 00:49:59,800 We we should talk a little bit. When you we did kind of shit 956 00:49:59,800 --> 00:50:01,720 talk that pick so. But so you guys were. 957 00:50:01,720 --> 00:50:02,720 Ready. We also should talk to other 958 00:50:02,720 --> 00:50:08,120 picks that I mean, I don't know what's going on there. 959 00:50:08,400 --> 00:50:11,920 It's an interesting question cuz they were the hottest company in 960 00:50:11,920 --> 00:50:14,080 the universe kind of when you picked them at the small scale, 961 00:50:14,080 --> 00:50:16,600 like they were like really, really, really yeah, Zeitgeist 962 00:50:16,840 --> 00:50:20,720 in their early stages and I don't know, they're sort of out 963 00:50:20,720 --> 00:50:21,920 there. People put their models up on 964 00:50:21,920 --> 00:50:25,600 Hugging Face, but maybe for some reason they have not gotten that 965 00:50:25,640 --> 00:50:27,560 AI premium like everyone else. Yeah. 966 00:50:27,560 --> 00:50:31,800 I mean, they did rate, I think they where, where were they at 967 00:50:31,840 --> 00:50:33,280 when I drafted them? I mean and they've. 968 00:50:33,680 --> 00:50:37,080 Weren't they at? Four, yeah, four, 4 billion, so 969 00:50:37,080 --> 00:50:39,640 4 1/2, yeah. So I don't know, I don't, it 970 00:50:39,640 --> 00:50:42,960 just seems like maybe hard to get another round done at a 971 00:50:42,960 --> 00:50:46,200 higher valuation there. So I think that I'm going to 972 00:50:46,200 --> 00:50:49,320 have to drop and free up a spot there. 973 00:50:51,480 --> 00:50:56,880 The other ones are a little trickier because they're all 974 00:50:56,880 --> 00:51:01,280 doing pretty well, but are they the next open? 975 00:51:01,400 --> 00:51:04,720 AI poolside is a black box. Yeah, poolside's a black. 976 00:51:04,720 --> 00:51:07,920 Poolside still feels in the bigger big good. 977 00:51:07,960 --> 00:51:10,800 Blood still hot, obviously Amjad's coming to the 978 00:51:10,800 --> 00:51:13,120 conference. XAI is great. 979 00:51:13,720 --> 00:51:19,680 Runway's pretty exciting in the video space 11 Labs is raising 980 00:51:19,680 --> 00:51:25,480 at seemingly every year at increasingly large valuations, 981 00:51:25,480 --> 00:51:27,520 so that's a good one. I think I'm just going to leave 982 00:51:27,520 --> 00:51:30,920 it at that and take 3 new 3 new picks. 983 00:51:31,160 --> 00:51:32,680 Great. All right, let's do this thing. 984 00:51:32,960 --> 00:51:37,040 All right, time to pick. Time to pick Max with the first 985 00:51:37,040 --> 00:51:41,280 pick of year 3. I'm going to go with my gut here 986 00:51:41,280 --> 00:51:43,600 even though this technically isn't the highest valuation on 987 00:51:43,600 --> 00:51:48,080 the board and just go with the strongest vibes I think 988 00:51:48,240 --> 00:51:54,000 available on the board and say cursor the coding application. 989 00:51:54,600 --> 00:51:56,440 You're locked into that. You're locked into that. 990 00:51:56,440 --> 00:51:57,920 I'm. Locked into cursor you wanted. 991 00:51:57,960 --> 00:52:01,480 That I think that's insane. I'm I'm overjoyed and I'm like, 992 00:52:01,520 --> 00:52:03,440 that was not one of the ones I was going to pick. 993 00:52:03,440 --> 00:52:05,480 I just think that. I just think that my other 994 00:52:05,480 --> 00:52:09,880 options are cursor competitors, essentially scale AI 995 00:52:09,880 --> 00:52:14,080 competitors, questionable foundational model companies. 996 00:52:14,160 --> 00:52:18,120 You know, I don't know, I just the heat on cursor is so hot 997 00:52:18,120 --> 00:52:20,440 right now so I'm happy to hear the negative case now that I 998 00:52:20,440 --> 00:52:22,560 picked it since Eric clearly has a strong. 999 00:52:22,560 --> 00:52:24,560 I just think it couldn't be hotter than it is. 1000 00:52:24,840 --> 00:52:26,760 Yeah, I mean, but that's what we said. 1001 00:52:26,760 --> 00:52:29,120 About but it's like 20 billion, right? 1002 00:52:29,120 --> 00:52:31,240 Why do we have? We have a listed here at, not. 1003 00:52:31,520 --> 00:52:33,400 Yeah, okay. I think it is 20, so maybe it is 1004 00:52:33,400 --> 00:52:34,520 number one right now. Yeah. 1005 00:52:35,040 --> 00:52:38,400 I mean, has there not been a lesson from this draft other 1006 00:52:38,400 --> 00:52:41,440 than a draft the highest valued company for. 1007 00:52:41,560 --> 00:52:44,480 The first pick, certainly, yeah, we're going for market caps, 1008 00:52:44,480 --> 00:52:47,680 not, not a sort of climb. So or two for two. 1009 00:52:48,880 --> 00:52:51,200 On the highest valued company being the best pick in 1010 00:52:51,200 --> 00:52:53,040 retrospect so far. So I don't know. 1011 00:52:53,160 --> 00:52:55,760 I can't. I can't, can't talk myself out 1012 00:52:55,760 --> 00:52:57,760 of taking. It, it is, it is rumoured at 1013 00:52:57,760 --> 00:53:00,520 around 30 billion. It's just not raised 30. 1014 00:53:00,520 --> 00:53:06,160 It's not raised at that level. So it's Max. 1015 00:53:06,160 --> 00:53:09,400 It's a good Max. Remember when the last two years 1016 00:53:09,400 --> 00:53:12,240 you've told us like that we were dumb for picking all the hypiest 1017 00:53:12,240 --> 00:53:14,640 companies and. No, I wanted xai. 1018 00:53:14,880 --> 00:53:18,680 I wanted xai, I wanted, I wanted XA clear, definitely. 1019 00:53:18,960 --> 00:53:21,120 Pull the tape. I I wanted xai. 1020 00:53:21,120 --> 00:53:22,760 So don't. Don't back. 1021 00:53:22,880 --> 00:53:24,040 Don't. Retcon that. 1022 00:53:24,320 --> 00:53:27,280 If Tesla and XAI ever merge, we're going to have like a big 1023 00:53:27,360 --> 00:53:30,120 challenge in evaluating that that I also believe. 1024 00:53:30,320 --> 00:53:35,640 That I Yeah, yeah, I believe I bid up Eric farther with open AI 1025 00:53:35,640 --> 00:53:36,880 than you did, James. So I was more. 1026 00:53:36,880 --> 00:53:38,280 Of a believer that's. True. 1027 00:53:38,280 --> 00:53:42,360 All right, so I'm next. With the next pick, Eric. 1028 00:53:44,400 --> 00:53:47,040 This is a hard one. Not so easy now, is it? 1029 00:53:47,040 --> 00:53:50,720 My go. It's hard to be the man in the 1030 00:53:50,720 --> 00:53:57,640 chair now. I am picking cognition labs. 1031 00:53:57,640 --> 00:53:58,200 OK. All right. 1032 00:53:58,720 --> 00:54:02,360 So you, so you literally. Spike for the Spike Pepsi Dick. 1033 00:54:02,360 --> 00:54:05,840 Or a joke. They're they're they're like you 1034 00:54:05,880 --> 00:54:08,760 bet What? You went to Pepsi, I picked Coke 1035 00:54:08,760 --> 00:54:10,600 and you picked Pepsi and you were shitting on. 1036 00:54:10,600 --> 00:54:13,280 My I'm betting on this is fucking Red Bull. 1037 00:54:13,280 --> 00:54:15,480 This is the next generation. They're like doing the new 1038 00:54:15,480 --> 00:54:17,480 thing. You guys are stuck in the pads 1039 00:54:17,520 --> 00:54:19,000 trying to sell sugar water, you know? 1040 00:54:19,080 --> 00:54:25,200 Like yeah, wait, how is Devin Red Bull? 1041 00:54:25,200 --> 00:54:28,320 Please explain. Just because it's like they've 1042 00:54:28,320 --> 00:54:30,520 got a new, they have a new growth strategy, they're doing 1043 00:54:30,520 --> 00:54:32,480 new things. I'm just saying it's like it's 1044 00:54:32,480 --> 00:54:34,160 not Pepsi. I mean, obviously, you know 1045 00:54:34,160 --> 00:54:36,920 what, Pepsi, I assume is bigger than Red Bull. 1046 00:54:36,920 --> 00:54:38,520 I don't actually know, but I'm just saying, you know, 1047 00:54:38,840 --> 00:54:41,200 technicians got some juice. They got new things going there. 1048 00:54:41,240 --> 00:54:43,320 They've, I feel like there's upside. 1049 00:54:43,360 --> 00:54:46,400 That's what I'm trying to say. They are the new, new coder on 1050 00:54:46,400 --> 00:54:47,720 the block. Even though they've been around, 1051 00:54:47,720 --> 00:54:51,320 it feels like, I feel like their momentum is is gonna be hitting 1052 00:54:51,640 --> 00:54:55,520 over the next 12 months. All right, sounds good. 1053 00:54:55,760 --> 00:54:56,960 Good pick. Good pick. 1054 00:55:00,280 --> 00:55:04,760 I have some tough choices to make here For my first pick. 1055 00:55:06,000 --> 00:55:08,560 You get 2 picks, I get 2 picks. Yeah, that actually helps a 1056 00:55:08,560 --> 00:55:10,280 little bit. We do a snake dress. 1057 00:55:10,280 --> 00:55:13,560 Yeah, yeah. I'm going to take for my first 1058 00:55:13,560 --> 00:55:18,520 pick Mercor which is the data labeling startup that. 1059 00:55:18,720 --> 00:55:21,920 That was my second choice. Isn't, isn't that also basically 1060 00:55:21,920 --> 00:55:23,800 a scale AI client? It is, yeah. 1061 00:55:23,800 --> 00:55:27,360 I'm OK, All right. Which did exit at $30 billion, 1062 00:55:27,360 --> 00:55:27,840 right? Get me wrong? 1063 00:55:27,880 --> 00:55:31,760 And and apparently when Windscale kind of got Aqua hired 1064 00:55:31,920 --> 00:55:35,240 by Meta, like a lot of business seems to have gone to Mercor 1065 00:55:35,240 --> 00:55:37,840 like because all the other labs were like, I don't want to send 1066 00:55:37,840 --> 00:55:42,000 all of our data to Meta. So just raised I think in 1067 00:55:42,000 --> 00:55:48,320 October, a $350 million round is my what my research says here. 1068 00:55:48,320 --> 00:55:51,480 And then valued at close to 10 billion. 1069 00:55:51,480 --> 00:55:55,320 So yeah, going with the the data labeling play, which is another 1070 00:55:55,320 --> 00:55:58,880 hot area of AI right now, definitely could come up as a 1071 00:55:58,880 --> 00:56:01,840 theme in the conference. I'm I'm happy either way with 1072 00:56:01,840 --> 00:56:04,160 what you go. I I think I'm I'm feeling you're 1073 00:56:04,160 --> 00:56:05,960 making my life easy for my next. Pick Really. 1074 00:56:06,120 --> 00:56:10,440 You're like but but. Because I don't have to make the 1075 00:56:10,440 --> 00:56:13,320 decision I'm saying. I think there are two obvious 1076 00:56:13,320 --> 00:56:14,920 ones to choose from. No. 1077 00:56:14,920 --> 00:56:17,920 And you're saying that I haven't choose one or the other, OK. 1078 00:56:18,120 --> 00:56:20,040 Yeah, and I don't have to. I'll just get the other one. 1079 00:56:21,280 --> 00:56:25,040 OK, well let me take a look here for my second pick. 1080 00:56:25,280 --> 00:56:29,320 I'm like, nervous. I take this very seriously. 1081 00:56:31,040 --> 00:56:33,680 I'm gonna have to go with thinking Machine Labs. 1082 00:56:33,680 --> 00:56:37,400 I just think it's another foundation model play and who 1083 00:56:37,400 --> 00:56:39,160 knows? Those are those are the only 1084 00:56:39,160 --> 00:56:43,840 ones that seem to matter in this draft in terms of potential 1085 00:56:43,840 --> 00:56:48,080 upside. Ride with and the insane acquire 1086 00:56:48,080 --> 00:56:49,720 potential there is. True, true. 1087 00:56:50,000 --> 00:56:52,720 And there seems to be a lot of true believers there because 1088 00:56:52,720 --> 00:56:57,800 like, it seems like Mark Zuck can't, can't poach anyone for 1089 00:56:58,080 --> 00:56:59,880 10s of billions. He got, he got 1. 1090 00:56:59,880 --> 00:57:01,200 He got, oh, he eventually got that guy. 1091 00:57:01,200 --> 00:57:03,200 Got the founder of. Thinking Oh yeah, yeah. 1092 00:57:03,440 --> 00:57:04,880 Yeah, that's right. Yeah, yeah, yeah. 1093 00:57:04,960 --> 00:57:07,920 No, he grabbed the founder of. Thinking fair enough. 1094 00:57:08,720 --> 00:57:11,320 The Yeah. And nobody knows what they're 1095 00:57:11,320 --> 00:57:14,320 doing, but yeah. They come out with interesting 1096 00:57:14,320 --> 00:57:17,040 research it seems like. Yeah. 1097 00:57:17,360 --> 00:57:20,680 But I agree it's a little bit of a hot just a Mira. 1098 00:57:21,160 --> 00:57:25,000 Murati, CTO of Open AI, founded this company. 1099 00:57:25,000 --> 00:57:27,160 Andreessen Horowitz is now a big backer. 1100 00:57:27,960 --> 00:57:31,000 But it's totally mysterious. And when this thing goes to 200 1101 00:57:31,000 --> 00:57:34,760 next year, you guys can't say that I got a steal cuz you had 1102 00:57:34,960 --> 00:57:36,680 the shot at this, you had the shot at this. 1103 00:57:36,680 --> 00:57:39,120 So I'm excited. I already have the ILIA 1. 1104 00:57:39,120 --> 00:57:42,240 So on one hand I was like, oh, safe super intelligence, have 1105 00:57:42,240 --> 00:57:45,040 thinking machines and have, I don't know, the pair of 1106 00:57:45,760 --> 00:57:48,680 speculative super hyped, you know, but. 1107 00:57:49,280 --> 00:57:52,480 It made it a little easier to pick right after Mercor, cuz I'm 1108 00:57:52,480 --> 00:57:55,840 like, all right, Mercor, we got real revenue and a real business 1109 00:57:55,840 --> 00:57:58,040 and something's happened in there. 1110 00:57:58,800 --> 00:58:01,200 But yeah, passing it back to Eric. 1111 00:58:01,200 --> 00:58:02,480 I'm back. It's back to me. 1112 00:58:02,760 --> 00:58:03,880 Yeah. All right. 1113 00:58:03,880 --> 00:58:05,680 This is a good one. I think you guys know what I'm 1114 00:58:05,680 --> 00:58:06,720 gonna pick, right? I mean. 1115 00:58:07,680 --> 00:58:09,600 I mean, I have a guess, yeah, but I don't want to this this. 1116 00:58:09,600 --> 00:58:11,160 Founder can eat the fucking world. 1117 00:58:11,160 --> 00:58:14,320 I sit down with him and I'm like what business aren't you going 1118 00:58:14,320 --> 00:58:19,440 to destroy Guillermo Co Versal I'm I'm picking up versal. 1119 00:58:19,440 --> 00:58:23,520 They first of all, even if they don't succeed in vibe coding 1120 00:58:23,520 --> 00:58:27,760 themselves, they are like doing a lot of the infrastructure work 1121 00:58:27,760 --> 00:58:30,760 for a lot of the vibe coding. So it's like lovable winds fine 1122 00:58:30,760 --> 00:58:34,280 like bill still use versal. So I I think Versal is like in a 1123 00:58:34,280 --> 00:58:39,280 cool position and I think Guillermo is a very talented CEO 1124 00:58:39,280 --> 00:58:42,160 and yeah, I think they're going, they're going place so happy to 1125 00:58:42,160 --> 00:58:46,240 pick up for a sell a sort of infrastructure play with, with, 1126 00:58:46,560 --> 00:58:49,760 you know, consumer aspects. Yeah, that definitely would have 1127 00:58:49,760 --> 00:58:53,640 been the next pick for me. Now it gets a lot more 1128 00:58:53,640 --> 00:58:57,440 interesting, I think, because I think there's a bit of a Cliff 1129 00:58:57,440 --> 00:59:01,240 here I get. Two, I'm lost now, yeah. 1130 00:59:01,400 --> 00:59:03,880 Yeah, I'm pretty lost, I mean. This is when it gets fun. 1131 00:59:04,120 --> 00:59:10,160 Yeah, I think thematically in the thinking machines category, 1132 00:59:10,200 --> 00:59:13,280 I feel like I, I just have to grab reflection AI here just 1133 00:59:13,280 --> 00:59:18,640 because open source foundational model company, it's American 1134 00:59:18,640 --> 00:59:20,800 apparently that's part of their story and we're going to have 1135 00:59:20,800 --> 00:59:24,360 some sort of national security defense by having another cool 1136 00:59:24,360 --> 00:59:27,240 foundational model company. I I don't know, it's got big 1137 00:59:27,240 --> 00:59:30,120 time backers, you know, Sequoia and DST and all these folks. 1138 00:59:30,120 --> 00:59:35,160 So, yeah, ultimately, I think just, I honestly know very, very 1139 00:59:35,160 --> 00:59:39,960 little about this, but yeah, I'm like, it's worth $8 billion and 1140 00:59:39,960 --> 00:59:43,880 it's foundational model company. Sign me up, Scotty. 1141 00:59:50,080 --> 00:59:54,160 Man, it's easy to be a VC. It's easy, especially when you 1142 00:59:54,160 --> 00:59:56,880 don't have to really pay. Like we're not like buy, you 1143 00:59:56,880 --> 00:59:58,360 know, we're not caring about the cost. 1144 00:59:58,360 --> 01:00:01,640 We don't worry about dilution. We only care about market cap. 1145 01:00:01,640 --> 01:00:05,240 So if you were in a VC some I think our incentives would be 1146 01:00:05,240 --> 01:00:06,000 different. I agree. 1147 01:00:06,000 --> 01:00:08,920 Like, OK, sure, I'm gonna get diluted like crime. 1148 01:00:08,920 --> 01:00:11,600 Do I want it at this price versus like am I? 1149 01:00:11,640 --> 01:00:14,120 Just trying. So the the way we set this up 1150 01:00:14,120 --> 01:00:17,600 incentivizes foundation models and it would have also 1151 01:00:17,600 --> 01:00:19,520 incentivized infrastructure companies. 1152 01:00:19,760 --> 01:00:22,920 Well, this is like this is a classic debate in fantasy 1153 01:00:22,920 --> 01:00:24,800 football, right? Is like you do snake draft or 1154 01:00:24,800 --> 01:00:27,720 you do auction draft where you actually have to like pay out of 1155 01:00:27,720 --> 01:00:29,400 your own budget and stuff like that. 1156 01:00:29,400 --> 01:00:33,400 So next, next, next time we can be a little bit more, a little 1157 01:00:33,400 --> 01:00:35,160 bit more savvy. Yeah. 1158 01:00:35,720 --> 01:00:38,440 OK. This next one I'm I'm actually 1159 01:00:38,440 --> 01:00:42,040 going to take what I consider to be a real risk on here because I 1160 01:00:42,040 --> 01:00:44,440 think we're starting to get into the risk part of the draft here. 1161 01:00:45,840 --> 01:00:49,600 Cerebra Valley alumnus here. I'm going to take lovable. 1162 01:00:49,600 --> 01:00:54,160 I just feel like the momentum around lovable as a brand name 1163 01:00:54,160 --> 01:00:57,000 or a meme or whatever the fuck you want to call it is just so 1164 01:00:57,000 --> 01:00:59,440 incredibly strong. And I don't necessarily think 1165 01:00:59,440 --> 01:01:02,000 the product is like that much better than a bunch of other 1166 01:01:02,000 --> 01:01:06,800 similar competitors or even ultimately cursor, you know, but 1167 01:01:06,920 --> 01:01:11,760 I think that it's just got that sort of like mimetic force and 1168 01:01:11,760 --> 01:01:14,240 growth. And I do think the founder is, 1169 01:01:14,320 --> 01:01:17,480 you know, has a lot of kind of narrative power around the. 1170 01:01:17,480 --> 01:01:20,000 Company you know I lit them up on Twitter they had the most 1171 01:01:20,000 --> 01:01:24,160 dark pattern unsubscribe no, I I don't and they're getting all 1172 01:01:24,160 --> 01:01:26,680 this shit on Twitter about like whether their growth metrics 1173 01:01:26,680 --> 01:01:29,040 make any sense I don't know no some red flags I was. 1174 01:01:29,040 --> 01:01:31,360 Giving them some of that shit. So maybe it's, you know, maybe 1175 01:01:31,360 --> 01:01:34,600 it's a dumb bet, but I do also think that like, probably there 1176 01:01:34,600 --> 01:01:38,800 will be a consumer brand in like make a website with an LLM and 1177 01:01:38,840 --> 01:01:41,560 maybe it will ultimately just be open AI and which point this is 1178 01:01:41,560 --> 01:01:43,680 all irrelevant. But they feel like they have by 1179 01:01:43,680 --> 01:01:47,640 far the strongest, like lead on being kind of the, you know, 1180 01:01:47,640 --> 01:01:51,360 consumer brand for vibe coding, for lack of a better term. 1181 01:01:51,960 --> 01:01:53,960 So yeah, I don't know, it's a bet. 1182 01:01:53,960 --> 01:01:57,160 I'm not in love with it, but I've got to take some shots. 1183 01:01:57,160 --> 01:01:59,840 And they feel like they have a very strong, you know, strong 1184 01:01:59,840 --> 01:02:01,640 memetic force behind them, I guess I would say. 1185 01:02:01,800 --> 01:02:04,120 And you're you're going to get a lot of picks That then is strap, 1186 01:02:04,120 --> 01:02:06,120 I think because Eric and I are really picking. 1187 01:02:06,600 --> 01:02:08,760 All right. I'm up and I can't give away all 1188 01:02:08,760 --> 01:02:10,400 the companies that I think are good left. 1189 01:02:10,440 --> 01:02:13,600 We're definitely deeper. It's less like, oh, just buy the 1190 01:02:13,600 --> 01:02:17,760 highest valuation now I'm I'm not willing to go for a there. 1191 01:02:17,880 --> 01:02:20,400 There's, you know, companies that I like sub 1 billion that 1192 01:02:20,400 --> 01:02:24,480 I'm like, Oh, you could bet on them, but I it seems risky. 1193 01:02:24,840 --> 01:02:28,360 I'm OK I'm embracing one of Max's themes, which is video is 1194 01:02:28,360 --> 01:02:32,680 doing better than voice. And I also do think, you know, 1195 01:02:32,720 --> 01:02:35,560 there it's more fun to this isn't like a fun business to 1196 01:02:35,560 --> 01:02:39,600 pick, but it's like selling to businesses is a good idea. 1197 01:02:39,600 --> 01:02:42,840 There's money to be made. That's sort of the glean case. 1198 01:02:43,160 --> 01:02:48,280 And so, and I really like the CEO of this company, Victor. 1199 01:02:48,520 --> 01:02:50,320 So I'm gonna pick Synthesia. Synthesia. 1200 01:02:50,360 --> 01:02:52,920 Yeah, okay, yeah. And it's also got the, it's 1201 01:02:52,920 --> 01:02:55,880 European undervalued sort of thing going. 1202 01:02:55,880 --> 01:02:59,160 And, and so yeah, I, I think, I think they'll go far. 1203 01:02:59,160 --> 01:03:02,160 I'm picking Synthesia. Well, they just leaked that they 1204 01:03:02,160 --> 01:03:04,680 turned on an acquisition offer for $3 billion, right? 1205 01:03:04,680 --> 01:03:06,000 I missed that. I didn't even know that. 1206 01:03:06,160 --> 01:03:07,840 Oh yeah, yeah. So they're definitely gonna 1207 01:03:07,840 --> 01:03:09,680 raise it like 5 or 6 pretty shortly. 1208 01:03:09,760 --> 01:03:12,800 Some great smart pick, smart pick, good value in the lower 1209 01:03:12,840 --> 01:03:16,880 part of the the draft if they're thinking about raising or you 1210 01:03:16,880 --> 01:03:19,040 know, close to 5, right. James. 1211 01:03:19,760 --> 01:03:22,000 James final pick. I'll pick. 1212 01:03:22,160 --> 01:03:27,120 I'm gonna go out on a limb as well here deep, deep in the 1213 01:03:27,120 --> 01:03:33,560 waivers pool right now, but in talks I believe to raise in the 1214 01:03:33,840 --> 01:03:39,000 2 billion range, Suno. Yeah, that was also pretty high 1215 01:03:39,040 --> 01:03:40,800 up there. I don't think that's deep in the 1216 01:03:40,800 --> 01:03:42,440 waivers is. It right, I mean right now if 1217 01:03:42,440 --> 01:03:43,760 you. They're they're, they're 1218 01:03:43,760 --> 01:03:46,360 obviously leaking all their ARR growth to close it around right 1219 01:03:46,360 --> 01:03:48,520 now. Yeah, I'm I'm just saying like 1220 01:03:48,840 --> 01:03:52,960 they their current valuation is sub a billion, you know, 500 1221 01:03:52,960 --> 01:03:53,800 million or something. Sure. 1222 01:03:54,000 --> 01:03:58,040 OK, so yeah, like yeah, there's been a lot of hype around them 1223 01:03:58,040 --> 01:04:01,560 in the last like week and they yeah, they leaked like 150 1224 01:04:01,560 --> 01:04:06,160 million or I don't know if they did, but they that leaked 150 1225 01:04:06,160 --> 01:04:08,840 million. Arr lot of discussion, obviously 1226 01:04:08,840 --> 01:04:12,720 a lot of potential IP risk, but every IP lawsuit these days 1227 01:04:12,720 --> 01:04:16,520 seems to be going in the way of the way of the the models. 1228 01:04:17,280 --> 01:04:21,880 So we'll see if that stands. But I think no matter what 1229 01:04:21,880 --> 01:04:24,640 they're really interesting acquisition or. 1230 01:04:25,200 --> 01:04:28,760 Spherical Valley alumni. Yes, I I I interviewed Michael 1231 01:04:28,760 --> 01:04:32,320 at Mikey at New York CVAI last year. 1232 01:04:32,920 --> 01:04:35,400 You looked into his eyes. And I was like, yeah, I want 1233 01:04:35,400 --> 01:04:39,640 this guy on my team. He's a, he's going places. 1234 01:04:40,880 --> 01:04:44,640 He was a yeah. OK, that's my pick. 1235 01:04:44,920 --> 01:04:45,480 Great. Nice. 1236 01:04:45,560 --> 01:04:47,120 All right, all. Right, James and I are done all. 1237 01:04:47,480 --> 01:04:49,800 Right, you guys are done. It's time to try to grab some 1238 01:04:49,800 --> 01:04:51,720 value off the. So how many do you have? 1239 01:04:51,960 --> 01:04:54,840 Off the ground ground here. Well I have 3 so far but I get 3 1240 01:04:54,840 --> 01:04:56,000 more. This, yeah. 1241 01:04:57,280 --> 01:04:59,400 Yeah, yeah, this is. Exciting. 1242 01:04:59,400 --> 01:05:00,760 There was some. This is a fun thing for you to 1243 01:05:00,760 --> 01:05:01,240 do. There are some. 1244 01:05:01,280 --> 01:05:02,960 This is my rebuilding opportunity here. 1245 01:05:03,160 --> 01:05:07,000 I did like Suno and Synthesia. They both would have been would. 1246 01:05:07,000 --> 01:05:08,520 Have been in the mix. Oh my gosh. 1247 01:05:08,720 --> 01:05:10,840 Yeah, Yeah. I literally have them in my 1248 01:05:10,880 --> 01:05:15,040 notepad ones, ones I wanted here for the first one. 1249 01:05:15,040 --> 01:05:17,760 I'm just going to grab one with a good valuation that we use a 1250 01:05:17,760 --> 01:05:22,200 lot at Volley because I think that I don't know, it's a solid 1251 01:05:22,200 --> 01:05:24,880 product and I have no idea if the business model is any good 1252 01:05:25,240 --> 01:05:29,400 or what the future holds here, but I'm going to I'm going to 1253 01:05:29,400 --> 01:05:32,720 take fall AIFALAI. Well, I've been hearing about 1254 01:05:32,720 --> 01:05:33,960 that one a lot. Interesting. 1255 01:05:34,040 --> 01:05:35,800 But it's not. I don't really know much about 1256 01:05:35,800 --> 01:05:38,440 it. They're essentially like a kind 1257 01:05:38,440 --> 01:05:44,040 of a clearing house or single API provider that lets you call 1258 01:05:44,040 --> 01:05:45,480 on any of the photo and video models. 1259 01:05:45,480 --> 01:05:47,960 So it kind of plays into this photo and photo and video theme, 1260 01:05:47,960 --> 01:05:50,600 which is like you're going to need to be able to switch 1261 01:05:50,600 --> 01:05:53,280 between these models really easily and grab the new ones. 1262 01:05:53,280 --> 01:05:55,240 And they take a little cut of that. 1263 01:05:55,240 --> 01:05:57,400 So it's almost like a marketplace for photo and video 1264 01:05:57,400 --> 01:06:00,440 models. And I don't know if that's 1265 01:06:00,480 --> 01:06:02,440 ultimately where value will accrue in the stack. 1266 01:06:02,440 --> 01:06:04,480 There's a lot of argument that their margins will be competed 1267 01:06:04,480 --> 01:06:08,640 to nothing or whatever. But like today, we at Volley as 1268 01:06:08,640 --> 01:06:12,040 a consumers company get a lot of value out of having like one 1269 01:06:12,040 --> 01:06:15,800 place to sort of call into for any kind of photo or image 1270 01:06:15,800 --> 01:06:19,200 generation. And so, yeah, I don't know, 1271 01:06:19,480 --> 01:06:22,920 feels like a. Invest in what you know, yeah. 1272 01:06:23,320 --> 01:06:24,840 Pretty highly valued at 4 billion as well. 1273 01:06:25,080 --> 01:06:27,720 Again, the opposite logic of normal investing, which is Oh 1274 01:06:27,760 --> 01:06:29,960 yeah, I want to take those high valuations. 1275 01:06:30,320 --> 01:06:32,960 Yeah, exactly right. I mean, there's this whole world 1276 01:06:32,960 --> 01:06:36,320 model startup theme that James and I are very familiar with, 1277 01:06:36,320 --> 01:06:39,960 which is this idea of like video, AI generated video that 1278 01:06:39,960 --> 01:06:43,800 you can walk around in and, you know, participate in as if it's 1279 01:06:43,800 --> 01:06:46,520 a video game. One of our good friends and 1280 01:06:46,520 --> 01:06:51,080 board member, Moritz buyer Lentz just raised $130 million for one 1281 01:06:51,080 --> 01:06:56,520 of these companies. But in this space, I'm going to 1282 01:06:56,520 --> 01:07:00,920 take, I think I'm gonna take World Labs, the Fei Fei Lee 1283 01:07:00,920 --> 01:07:04,280 startup. I think that it's just too much 1284 01:07:04,280 --> 01:07:11,400 name brand to avoid. I mean, I've seen D cart, I 1285 01:07:11,400 --> 01:07:13,080 haven't seen general intuitions product. 1286 01:07:13,320 --> 01:07:15,920 I ultimately think Demosus, Abbas and Jeannie will probably 1287 01:07:15,920 --> 01:07:18,800 win this whole thing or whatever, but maybe this is a 1288 01:07:18,800 --> 01:07:20,280 good acquisition or role up play at some. 1289 01:07:20,280 --> 01:07:21,720 Point what? What did you think about 1290 01:07:21,720 --> 01:07:23,880 Descartes? Why not pick Descartes over 1291 01:07:23,880 --> 01:07:27,160 World Labs? Yeah, I literally am just doing 1292 01:07:27,160 --> 01:07:31,680 this based on like name brand. And my assumption is that she 1293 01:07:31,680 --> 01:07:34,840 can raise a shit load of money and that might ultimately lead 1294 01:07:34,840 --> 01:07:37,280 to them winning the compute and the data war here. 1295 01:07:37,280 --> 01:07:39,880 Yeah, but obviously that could be totally wrong. 1296 01:07:40,520 --> 01:07:45,280 OK, last pick. I think the last one I'm gonna 1297 01:07:45,280 --> 01:07:49,720 take, and this is, again, probably just biased on what I 1298 01:07:49,720 --> 01:07:56,160 know, is I'm gonna grab Sesame, which is this glasses startup 1299 01:07:56,160 --> 01:07:59,400 that's focused on realistic voice assistance built into 1300 01:07:59,400 --> 01:08:03,160 smart glasses, which sounds terrible, terrible in a lot of 1301 01:08:03,160 --> 01:08:03,840 ways. I. 1302 01:08:03,920 --> 01:08:07,280 Would never pick that. They just raised $250 million 1303 01:08:07,280 --> 01:08:11,480 from Sequoia Capital, not nothing. 1304 01:08:11,480 --> 01:08:14,120 And I do think their TTS voice is quite good. 1305 01:08:14,120 --> 01:08:18,960 And I think that ultimately, it's interesting to see people 1306 01:08:18,960 --> 01:08:21,479 playing in the sort of voice and glasses space. 1307 01:08:22,359 --> 01:08:26,960 And they might get rolled up by either Mark Zuckerberg or Apple 1308 01:08:26,960 --> 01:08:29,479 or someone else like that at some point. 1309 01:08:29,479 --> 01:08:34,520 So it feels like just a, a real, a real Yolo play here. 1310 01:08:35,560 --> 01:08:38,520 And yeah, I don't know, it's just kind of fun and interesting 1311 01:08:38,520 --> 01:08:41,319 and I think the voice product is pretty good, so I feel like why 1312 01:08:41,319 --> 01:08:44,120 not invest in that? So all right, that's it. 1313 01:08:44,359 --> 01:08:46,359 Nice. Do you guys want to do it can. 1314 01:08:46,359 --> 01:08:47,760 We review. Quick review. 1315 01:08:48,399 --> 01:08:51,080 Sure. Yes, I will kick it off my 1316 01:08:51,080 --> 01:08:54,040 roster. Previously Databricks, Cohere, 1317 01:08:54,720 --> 01:08:57,760 Sierra and that was it. And I added a bunch of new 1318 01:08:57,760 --> 01:09:00,520 companies today. Cursor, Star of the Show, 1319 01:09:00,760 --> 01:09:04,720 Reflection, AI, Lovable, Fall, World Labs, and Sesame. 1320 01:09:06,000 --> 01:09:10,319 And Eric? I had Open AI, Glean, Mistral, 1321 01:09:10,319 --> 01:09:14,640 Perplexity, Safe, Super Intelligence, and Harvey and I 1322 01:09:14,640 --> 01:09:18,600 picked up Cognition, Versal, and Synthesia. 1323 01:09:19,279 --> 01:09:24,880 And I had anthropic replit XAI runway 11 labs poolside. 1324 01:09:25,040 --> 01:09:29,640 I just added Mercor, Thinking Machine Labs and SUNO. 1325 01:09:30,359 --> 01:09:31,399 Good drafts. Yeah, fun. 1326 01:09:31,520 --> 01:09:34,040 We have one more. Episode before St. 1327 01:09:34,040 --> 01:09:38,000 Roll Valley on November 12th, we'll be making predictions or 1328 01:09:38,000 --> 01:09:43,920 reacting to some AI predictions. And yeah, if you're in San 1329 01:09:43,920 --> 01:09:48,200 Francisco and startup founder or investor, reach out about 1330 01:09:48,439 --> 01:09:49,920 Cerebral Valley on November 12th. 1331 01:09:50,319 --> 01:09:53,000 Thank you for tuning into this week's episode of the podcast. 1332 01:09:53,000 --> 01:09:55,160 If you're new here, please like and subscribe. 1333 01:09:55,160 --> 01:09:57,840 It really helps the channel. We're building a YouTube 1334 01:09:57,840 --> 01:09:58,960 channel. I think you can tell we're 1335 01:09:58,960 --> 01:10:01,680 investing a lot more in our production and we appreciate 1336 01:10:01,680 --> 01:10:05,840 your support. And if you want the data insider 1337 01:10:05,840 --> 01:10:10,280 takes real reporting, go to newcomer.co and subscribe to the 1338 01:10:10,280 --> 01:10:12,880 sub stack as well. Thanks for following along.