1 00:00:00,040 --> 00:00:03,720 Tom published his first post Tom Newcomer byline How does it 2 00:00:03,720 --> 00:00:05,760 feel? Feels like I'm born again. 3 00:00:08,320 --> 00:00:11,240 One of our mutual friends, I think Corey, the information was 4 00:00:11,240 --> 00:00:13,240 like, I'm going to call it Doe Tom whenever you're right here, 5 00:00:13,240 --> 00:00:16,239 it's like Tom a newcomer. I don't know it's. 6 00:00:16,400 --> 00:00:18,360 Funny, yeah. We haven't discussed branding on 7 00:00:18,360 --> 00:00:20,720 my stories, but if we could have some sort of like a hover 8 00:00:20,720 --> 00:00:24,680 graphic or something, we're like if the mouse goes over newcomer, 9 00:00:24,680 --> 00:00:26,480 like my name kind of fades in, flashes in. 10 00:00:27,280 --> 00:00:29,320 Yeah, let's talk about it. I'm willing to work it into comp 11 00:00:29,320 --> 00:00:32,520 discussion. This episode is presented by 12 00:00:32,520 --> 00:00:36,440 Breck's The Financial Stack that founders and VCs can truly bank 13 00:00:36,440 --> 00:00:38,480 on. Imagine what your founders could 14 00:00:38,480 --> 00:00:41,720 do with their runway if they had a banking solution that had no 15 00:00:41,720 --> 00:00:46,520 minimums, no transaction fees, and 20 times the standard FDIC 16 00:00:46,520 --> 00:00:49,160 protection. Plus they could earn an industry 17 00:00:49,160 --> 00:00:52,360 leading yield while maintaining access to funds whenever needed. 18 00:00:52,800 --> 00:00:56,040 Brex simplifies financial services for start-ups so they 19 00:00:56,040 --> 00:00:59,080 can focus on building. Connect your portfolio to the 20 00:00:59,080 --> 00:01:01,560 financial stack that one in three US venture backed 21 00:01:01,560 --> 00:01:07,760 start-ups already use. Check out brex.com/banking Dash 22 00:01:07,920 --> 00:01:12,240 solutions. Your piece was basically markets 23 00:01:12,240 --> 00:01:15,520 collapsing, but otherwise it would have been a great business 24 00:01:15,520 --> 00:01:19,000 news period for the foundation model companies, which is not 25 00:01:19,000 --> 00:01:20,960 your disposition. I feel like almost like I didn't 26 00:01:21,000 --> 00:01:23,960 order you to come in and say like tell everybody how AI is 27 00:01:23,960 --> 00:01:25,400 doing great. It's like you would be the guy 28 00:01:25,400 --> 00:01:27,880 to like, be like, I don't know, I'm worried about everything. 29 00:01:28,280 --> 00:01:31,960 Eric's hitting the techno optimism whip like you got to 30 00:01:31,960 --> 00:01:33,760 publish a positive story. Right. 31 00:01:33,760 --> 00:01:36,080 I wish you'd come in negative. So it didn't sound like, oh, 32 00:01:36,080 --> 00:01:39,080 you're writing for us, the AI conference people, You're going 33 00:01:39,160 --> 00:01:42,880 to give a booster take What? What convinced you that like 34 00:01:42,880 --> 00:01:45,080 things were going well in Foundation model? 35 00:01:45,080 --> 00:01:46,960 Work. Yeah, I really wanted to come in 36 00:01:46,960 --> 00:01:49,880 with a negative piece and then quit right after that and say, 37 00:01:49,920 --> 00:01:53,680 like Eric rejected my my contrarian take, my contrarian 38 00:01:53,680 --> 00:01:54,920 market take. Spiked. 39 00:01:54,960 --> 00:01:57,400 Yeah, exactly. Immediately, like this is who 40 00:01:57,400 --> 00:02:00,400 pays the bills here at Newcomer. No, On the contrary, Eric. 41 00:02:00,560 --> 00:02:02,360 He wants harder hitting stuff than what I'm delivering. 42 00:02:02,360 --> 00:02:06,400 I'm actually very worried. I was talking to some people and 43 00:02:06,400 --> 00:02:10,680 I just wanted to get, you know, a dipstick test of where AI was 44 00:02:10,680 --> 00:02:14,360 at the moment. And much to my chagrin, much to 45 00:02:14,360 --> 00:02:17,280 my disappointment, it was a very strong start of the year. 46 00:02:17,680 --> 00:02:19,760 You know, like I, I have this quote in the story, but I was 47 00:02:19,760 --> 00:02:23,480 talking to an investor who was telling me we all had in mind 48 00:02:23,480 --> 00:02:26,200 this idea that competition was the most important thing and 49 00:02:26,200 --> 00:02:28,800 there were only going to be a small number of winners and we 50 00:02:28,800 --> 00:02:30,920 needed to back those winners. And, and everyone else is going 51 00:02:30,920 --> 00:02:34,040 to be shit out of luck. And at least so far, it turns 52 00:02:34,040 --> 00:02:37,560 out that like the market size is large and there are enough 53 00:02:37,560 --> 00:02:40,960 people on the consumer and business side who want to pay 54 00:02:40,960 --> 00:02:43,600 for this stuff that the businesses, you know, and we 55 00:02:43,600 --> 00:02:46,720 were looking mostly at foundation model companies. 56 00:02:46,720 --> 00:02:49,480 So the ones that have raised the most money, but are building the 57 00:02:49,480 --> 00:02:52,320 tech that's powering almost all the applications. 58 00:02:53,200 --> 00:02:56,800 They had really good quarters and their business grew 59 00:02:57,160 --> 00:02:59,600 substantially. And, you know, in certain cases 60 00:02:59,600 --> 00:03:03,000 above initial projections, basically things are going very 61 00:03:03,000 --> 00:03:05,440 well in AI right now. And the only thing that could 62 00:03:05,440 --> 00:03:08,240 maybe derail it in the short term is the economy collapsing 63 00:03:08,400 --> 00:03:11,400 is the markets, you know, shitting the bed and, and 64 00:03:11,400 --> 00:03:13,640 everyone realizing that they're going to pull back on their 65 00:03:13,640 --> 00:03:15,880 spending because right now at least they're buying into the 66 00:03:15,880 --> 00:03:18,680 hype. And and so, yeah, that's the 67 00:03:18,680 --> 00:03:20,680 backdrop of it. It's like we've got all these 68 00:03:20,680 --> 00:03:24,000 numbers, we have these exclusive figures on, on how all the 69 00:03:24,000 --> 00:03:25,840 foundation model companies are doing, yeah. 70 00:03:25,920 --> 00:03:29,120 Give us a give us a tier ranking like who's who's doing the best 71 00:03:29,160 --> 00:03:31,680 of these foundation model companies right now for your for 72 00:03:31,680 --> 00:03:34,240 your numbers. Good, good week, bad week, good 73 00:03:34,240 --> 00:03:37,040 good quarter, bad quarter. It's kind of hard to dislodge 74 00:03:37,040 --> 00:03:40,200 open AI right now. They're really, they're really 75 00:03:40,200 --> 00:03:43,520 killing it as much as it kills me 'cause it's a much better 76 00:03:43,520 --> 00:03:46,320 story when they're not. I'm sorry, but you know, 77 00:03:46,320 --> 00:03:49,000 relentless product roll out pace, Yeah. 78 00:03:49,000 --> 00:03:51,760 We're gonna talk about O3, which I'm extremely excited about. 79 00:03:51,760 --> 00:03:54,880 We'll talk about the actual tech after we talk about sort of the 80 00:03:54,880 --> 00:03:56,680 business, you know, revenue. Right. 81 00:03:56,680 --> 00:03:59,400 So, so the the product stuff which we'll get to the revenue 82 00:03:59,400 --> 00:04:02,560 stuff which they are projecting that they could get as much as 83 00:04:02,560 --> 00:04:06,040 12, you know, almost 13 billion in revenue this year and that 84 00:04:06,040 --> 00:04:09,840 seems very plausible based on their user growth. 85 00:04:10,000 --> 00:04:12,440 You know, Sam had this. I hadn't watched it until I was 86 00:04:12,440 --> 00:04:14,880 doing the story. Very funny interview at the Ted 87 00:04:14,880 --> 00:04:20,079 conference where he kind of let slip that their users doubled in 88 00:04:20,079 --> 00:04:23,080 the last couple of weeks and, you know, back of the envelope 89 00:04:23,080 --> 00:04:23,520 you could. Thanks. 90 00:04:23,520 --> 00:04:27,320 Studio Ghibli, right? I mean, it's got to be mostly 91 00:04:27,400 --> 00:04:29,560 right. Like I that's got to be a lot of 92 00:04:29,560 --> 00:04:31,360 it. I feel like the yeah, the image 93 00:04:31,360 --> 00:04:35,000 generation has to just be overwhelming demand coming from. 94 00:04:35,000 --> 00:04:37,560 That nothing like a product. People actually like that. 95 00:04:37,560 --> 00:04:39,600 It's like, oh, let me show off using it to everybody. 96 00:04:39,600 --> 00:04:42,680 It's not like a social feature. It's like literally something so 97 00:04:42,680 --> 00:04:45,920 cool that it becomes like marketing for your product. 98 00:04:46,480 --> 00:04:48,200 But your Anthropic was doing well too. 99 00:04:48,200 --> 00:04:51,080 It is not just open AI. Right, Yep, Anthropic, which we 100 00:04:51,080 --> 00:04:55,040 really dove into their numbers, they finished off 24 with like 101 00:04:55,400 --> 00:04:58,040 950,000,000 in revenue. So very solid year. 102 00:04:58,320 --> 00:05:00,320 And you know, already at the beginning of the year they're 103 00:05:00,320 --> 00:05:04,240 tracking to do annualized 2 billion, well over 2 billion. 104 00:05:04,280 --> 00:05:06,720 And by the end of the year, if growth continues to pace, it 105 00:05:06,720 --> 00:05:09,400 could be as much as 4 billion. So that's still, you know, 1/3 106 00:05:09,640 --> 00:05:12,400 of open AI business, which is interesting, you know, and I 107 00:05:12,400 --> 00:05:14,920 think we're paying attention to the fact that one is like head 108 00:05:14,920 --> 00:05:17,600 and shoulders above the rest. But like Anthropic is really 109 00:05:17,600 --> 00:05:22,680 established itself as the enterprise AI model maker of 110 00:05:22,680 --> 00:05:25,320 choice. And if when it comes to coding, 111 00:05:25,320 --> 00:05:28,280 Claude is sort of where it's at. What's interesting with that is 112 00:05:28,280 --> 00:05:30,600 as I was doing this piece, like literally we were going to, you 113 00:05:30,600 --> 00:05:33,200 know, getting close to to putting the last touches on it. 114 00:05:33,560 --> 00:05:37,960 News breaks that open AIS and talks to acquire windsurf, which 115 00:05:37,960 --> 00:05:40,000 is this big AI coding. What big? 116 00:05:40,000 --> 00:05:42,520 But they're like, you know, a, a somewhat positive. 117 00:05:42,560 --> 00:05:44,600 Codium until like a couple days ago. 118 00:05:44,680 --> 00:05:46,720 Yeah, I didn't even know about windsurf, to be honest. 119 00:05:47,400 --> 00:05:50,760 Windsurf was the product you know they can be with cursor 120 00:05:50,760 --> 00:05:53,320 which is called any sphere. But for some reason, this 121 00:05:53,320 --> 00:05:55,680 category is full of companies with products that are more 122 00:05:55,680 --> 00:05:58,120 famous than the underlying companies Anyway. 123 00:05:58,400 --> 00:06:02,360 Yeah, Open AI is supposed to be buying it for 3 billion, which, 124 00:06:03,360 --> 00:06:06,000 you know, will scare the shit out of every application 125 00:06:06,000 --> 00:06:06,600 company. Or what do? 126 00:06:06,720 --> 00:06:08,520 Yeah. What do you guys make of that? 127 00:06:08,760 --> 00:06:11,040 I mean my first take on it 'cause I actually did not 128 00:06:11,040 --> 00:06:13,720 realize this until I was, you know, digging into Anthropics 129 00:06:13,720 --> 00:06:16,040 numbers. But those coding application 130 00:06:16,040 --> 00:06:18,400 start-ups, the any, What the fuck is it? 131 00:06:18,400 --> 00:06:19,880 Any sphere? Any sphere? 132 00:06:20,160 --> 00:06:21,320 They should. Just change that. 133 00:06:21,320 --> 00:06:23,960 Everyone calls it cursor and podium anyway guys, I I think 134 00:06:23,960 --> 00:06:27,720 generally across. AI so there is a company called 135 00:06:27,840 --> 00:06:31,080 Any Scale and you know, I'm like, I do all this conference 136 00:06:31,080 --> 00:06:33,880 organizing and like we we did this enterprise 30 list. 137 00:06:34,120 --> 00:06:39,000 So I accidentally emailed like any scale about any sphere and 138 00:06:39,000 --> 00:06:41,480 like they both ended up on the same thread. 139 00:06:41,680 --> 00:06:45,560 So I was on this like comical thread where like both of them 140 00:06:45,560 --> 00:06:47,800 were replying and I wanted one of them. 141 00:06:47,960 --> 00:06:51,120 I ended up talking to any scale just because it's like, oh, now 142 00:06:51,120 --> 00:06:52,680 we're on a thread. You're another interesting 143 00:06:52,680 --> 00:06:54,280 company. Like let's do a call. 144 00:06:54,560 --> 00:06:56,760 But yeah, it's extremely confusing. 145 00:06:57,120 --> 00:06:58,720 That's that. That's your growth pitch to 146 00:06:58,720 --> 00:07:00,360 these people. It's like since we're on the 147 00:07:00,360 --> 00:07:02,560 phone. Yeah, since I mistook. 148 00:07:02,560 --> 00:07:04,760 You for another company. Way it's like now that you've 149 00:07:04,760 --> 00:07:07,800 wasted our time, like you at least need to hear our spiel, 150 00:07:07,840 --> 00:07:10,040 you know, like. Anyway, we should have a whole 151 00:07:10,040 --> 00:07:12,440 episode on the shitty naming across all of AI. 152 00:07:12,440 --> 00:07:15,800 Like why O 4 is the less powerful to O3? 153 00:07:16,880 --> 00:07:19,640 Absolutely, yeah. I would say I feel like 154 00:07:19,680 --> 00:07:22,800 honestly, any sphere, any scale has become like a recurring 155 00:07:22,800 --> 00:07:25,040 character on this podcast to this point. 156 00:07:25,040 --> 00:07:28,280 Just like pick a name. Yeah, any, any port in the 157 00:07:28,280 --> 00:07:30,960 storm. So their acquisition of them is 158 00:07:30,960 --> 00:07:33,640 actually could be hugely problematic for Anthropic 159 00:07:33,640 --> 00:07:37,200 because one of the, you know, significant drivers of business 160 00:07:37,200 --> 00:07:40,040 for Anthropic are these coding startups that incorporate 161 00:07:40,040 --> 00:07:43,880 Anthropic's models into their application and that that powers 162 00:07:43,880 --> 00:07:46,480 the whole thing. And you got to assume if open 163 00:07:46,520 --> 00:07:50,920 AI, you know, continue like, like, like gets this acquisition 164 00:07:50,920 --> 00:07:53,880 across the line, they're going to cut out all of Anthropic's 165 00:07:53,880 --> 00:07:56,680 models. And I don't know exactly how 166 00:07:56,680 --> 00:07:59,520 much revenue I should look into that one, but it's certainly an 167 00:07:59,520 --> 00:08:02,400 offensive move on, on on open a is part. 168 00:08:02,400 --> 00:08:05,120 So that's something that to to watch that deal aside, 169 00:08:05,200 --> 00:08:07,480 Anthropic's in a great place and there's a huge amount of 170 00:08:07,480 --> 00:08:10,880 confidence there. And again, for the second place 171 00:08:11,320 --> 00:08:14,080 player to be in a good space generally means good things for 172 00:08:14,080 --> 00:08:16,840 the health of the sector. I I don't think it's just that 173 00:08:16,840 --> 00:08:19,080 Anthropic could lose the revenue. 174 00:08:19,120 --> 00:08:24,520 I think what matters is that Open AI is going further into 175 00:08:24,520 --> 00:08:27,360 the application space, right? It's not just chat. 176 00:08:27,360 --> 00:08:30,800 ChatGPT is their application. It would seem they want to 177 00:08:30,800 --> 00:08:34,280 provide other applications, which means anyone built on top 178 00:08:34,280 --> 00:08:36,320 of Open AI has their ears justified. 179 00:08:36,919 --> 00:08:40,280 I can reveal here for the first time that one sentence in Tom's 180 00:08:40,280 --> 00:08:43,120 story was written entirely by O3. 181 00:08:43,240 --> 00:08:46,120 You can pause here to go see if you can identify it. 182 00:08:46,360 --> 00:08:50,960 But it it was any AI app acquisition will jolt the 183 00:08:50,960 --> 00:08:53,920 remaining independence, confirming their fear that Chat 184 00:08:53,920 --> 00:08:57,600 GPT's parent won't settle for being just the brains behind 185 00:08:57,600 --> 00:09:02,000 everyone else's software. I had the one convoluted 186 00:09:02,000 --> 00:09:04,600 sentence and it rewrote it and you're like ChatGPT is better. 187 00:09:05,120 --> 00:09:07,600 I want to let listeners in. OK, You kind of got to behind 188 00:09:07,600 --> 00:09:09,520 the scenes here. This is a line that Eric wrote, 189 00:09:10,160 --> 00:09:12,120 not a line that I wrote. Because. 190 00:09:12,680 --> 00:09:16,160 Anytime anytime I put my language into ChatGPT and I say 191 00:09:16,160 --> 00:09:18,600 can you make this sound better? They say I can't. 192 00:09:19,160 --> 00:09:22,480 I truly can't. This is this is peak human 193 00:09:22,480 --> 00:09:25,840 intelligence, and it would be foolhardy for me to suggest 194 00:09:25,840 --> 00:09:27,920 anything. It just shuts off. 195 00:09:28,040 --> 00:09:30,400 When you try, yeah, I actually get a call from Brad Lightcap 196 00:09:30,400 --> 00:09:33,800 every time I put my wording in there and say we'd actually like 197 00:09:33,800 --> 00:09:38,360 to refund you $20, Sir. So it's the program. $200 now, 198 00:09:38,440 --> 00:09:39,880 yeah. Yeah, this stuff's getting 199 00:09:39,880 --> 00:09:42,560 better. I, I think the, the market right 200 00:09:42,560 --> 00:09:45,760 now is, is really is really strong and it pains me to say 201 00:09:45,760 --> 00:09:47,200 it. Well, what about, I mean, even 202 00:09:47,200 --> 00:09:49,840 in kind of apps you you looked into perplexity too, right? 203 00:09:49,840 --> 00:09:52,560 Like how are they doing? Briefly, we did, yeah. 204 00:09:53,080 --> 00:09:54,520 Good. You're doing good. 205 00:09:54,640 --> 00:09:57,040 Good. Yeah, I mean the CEO of 206 00:09:57,040 --> 00:10:00,320 Perplexity, you know, Arvind Srinivas, he's out there a lot. 207 00:10:00,320 --> 00:10:03,800 He likes promoting the company. And so he he tweeted out that, 208 00:10:03,840 --> 00:10:07,920 you know, they had reached 100 million in ARR in March. 209 00:10:07,920 --> 00:10:10,920 But digging a little bit deeper into them, I found that, you 210 00:10:10,920 --> 00:10:13,240 know, there's from the investor side of things, they're 211 00:10:13,240 --> 00:10:16,520 expecting that to almost double by the end of the year, so they 212 00:10:16,520 --> 00:10:19,320 feel really strong about stuff. You talked about opening eyes a 213 00:10:19,320 --> 00:10:22,320 big player. I mean, Google's Gemini 2.5 is 214 00:10:22,320 --> 00:10:23,720 hitting a bunch of benchmarks too. 215 00:10:23,720 --> 00:10:27,960 But meta seems like kind of fell off with Llama, like it's been 216 00:10:27,960 --> 00:10:30,440 doing nearly as well. I know obviously there's reports 217 00:10:30,440 --> 00:10:33,880 of engineers taking that they worked on the latest model off 218 00:10:33,880 --> 00:10:36,720 their LinkedIn or clarifying that they did not work on the 219 00:10:36,720 --> 00:10:39,640 latest model on their LinkedIn because it's just not. 220 00:10:39,680 --> 00:10:43,120 Doing the The accusation is that what Llama for was built to do 221 00:10:43,120 --> 00:10:46,560 well on the test, but not necessarily that big of an 222 00:10:46,560 --> 00:10:47,960 improvement. Yeah. 223 00:10:47,960 --> 00:10:50,040 I mean, first of all, I relate. There are certain jobs that I've 224 00:10:50,040 --> 00:10:52,160 had that I've taken off of my LinkedIn as well as I. 225 00:10:52,680 --> 00:10:56,080 Never know. So it's nice to know that, you 226 00:10:56,080 --> 00:10:58,240 know, you know, AI researchers are just like us. 227 00:10:59,280 --> 00:11:02,360 Llama's a kind of huge question mark right now. 228 00:11:02,520 --> 00:11:05,320 People were very impressed with the earlier iterations of it. 229 00:11:05,320 --> 00:11:08,240 I mean, this was like a high quality open, you know, or free 230 00:11:08,240 --> 00:11:11,840 to use, you know, thing that basically became the standard 231 00:11:11,840 --> 00:11:14,760 open source model. And then Deep sea came along 232 00:11:14,800 --> 00:11:17,160 and, you know, had deep reasoning capabilities, which I 233 00:11:17,160 --> 00:11:18,800 think really appeal to a lot of people. 234 00:11:18,800 --> 00:11:21,800 And then Llama 4 was supposed to come out and blow everything out 235 00:11:21,800 --> 00:11:25,520 of the water and they had, like you were mentioning Madeline. 236 00:11:26,000 --> 00:11:28,480 You know, these benchmarks look very impressive, but when people 237 00:11:28,480 --> 00:11:31,200 were using it, like when developers actually were trying 238 00:11:31,200 --> 00:11:34,560 to apply it, they were like, man, this sucks. 239 00:11:34,680 --> 00:11:38,000 You know, it's like very verbose and it's answers are much worse 240 00:11:38,000 --> 00:11:40,040 than, you know, what the competitive ones are, are 241 00:11:40,040 --> 00:11:43,240 offering. And it got to a point where AVP 242 00:11:43,240 --> 00:11:47,280 of AI at Meta had to get out there and say like, well, we're 243 00:11:47,280 --> 00:11:49,040 working through some bugs right now. 244 00:11:49,040 --> 00:11:51,080 We're trying to make sure that everything works for our for our 245 00:11:51,080 --> 00:11:53,360 users. And these accusations that we 246 00:11:53,360 --> 00:11:56,560 were training Llama on the test set so that it could perform 247 00:11:56,560 --> 00:11:58,920 well on the benchmarks is totally unfounded. 248 00:11:59,440 --> 00:12:03,000 And look, I don't know yet and I'll look into it as to, you 249 00:12:03,000 --> 00:12:05,080 know, what true or isn't true. But once you're doing that kind 250 00:12:05,080 --> 00:12:08,480 of damage control, you kind of already lost like it's a it's 251 00:12:08,480 --> 00:12:11,760 APR game to a degree. And so much of open source is 252 00:12:11,760 --> 00:12:13,760 like, hey, we're cool in the space. 253 00:12:13,760 --> 00:12:15,360 We're going to attract engineers. 254 00:12:15,600 --> 00:12:18,240 I mean, to some degree, Mehta's just trying to kneecap everybody 255 00:12:18,240 --> 00:12:20,000 else and be like, we give it away for free. 256 00:12:20,000 --> 00:12:22,080 You know, there are fewer giant tech companies. 257 00:12:22,080 --> 00:12:25,760 And if DeepSeek is playing that role, you know, Mehta still 258 00:12:25,800 --> 00:12:28,120 wins. And sort of making it hard for 259 00:12:28,120 --> 00:12:31,360 for everybody else to compete. But yeah, llamas need to, you 260 00:12:31,360 --> 00:12:34,000 know, need to rebound, I think on the next one. 261 00:12:34,560 --> 00:12:37,400 And I'm kind of, I'm, I'm sympathetic to them because 262 00:12:37,560 --> 00:12:40,560 these models like you do these training runs and you, you're 263 00:12:40,560 --> 00:12:43,200 kind of get what you get. And if it's not good enough, you 264 00:12:43,200 --> 00:12:46,800 do another one And, and either you release it or you don't. 265 00:12:46,800 --> 00:12:49,960 And they're clearly, I mean, if it truly is as bad as some 266 00:12:49,960 --> 00:12:51,720 people are saying, you know, there was a lot of pressure on 267 00:12:51,720 --> 00:12:54,920 them to just release something and, you know, they put out 268 00:12:54,920 --> 00:12:56,720 something that maybe wasn't very good and like. 269 00:12:56,720 --> 00:12:58,880 That's the beauty of everybody's competing. 270 00:12:58,880 --> 00:13:01,520 We should be happy, you know. I do feel like separate from all 271 00:13:01,520 --> 00:13:04,920 of these numbers on how these companies are doing, looking at 272 00:13:04,920 --> 00:13:09,040 the technical specs, like people are really excited about 03. 273 00:13:09,040 --> 00:13:11,600 It's like some people are saying we're hitting the AGI moment 274 00:13:11,600 --> 00:13:14,520 right now. Eric, you were AGI momenting all 275 00:13:14,520 --> 00:13:17,840 over Twitter this week. You're like, oh, yeah, you. 276 00:13:17,880 --> 00:13:20,400 I was up. I was declaring on Twitter like, 277 00:13:20,400 --> 00:13:23,760 oh, history will reflect. This is a huge moment, but the 278 00:13:23,760 --> 00:13:25,520 headlines are about everything. But you were. 279 00:13:25,560 --> 00:13:28,920 In our slack last night at like 9:00 PM just being like wow, how 280 00:13:28,920 --> 00:13:30,320 did it do this? You were. 281 00:13:30,320 --> 00:13:33,120 Like I told you guys, you're not obligated to engage with me when 282 00:13:33,120 --> 00:13:37,080 I'm just like slacking away. But Tyler Cowen said AGI, 283 00:13:37,080 --> 00:13:39,640 seriously, he really thinks his AGI. 284 00:13:39,840 --> 00:13:44,000 I was getting some pushback from AI diehards that are like, OK, 285 00:13:44,000 --> 00:13:47,840 O3 isn't so much better than Gemini and Anthropic, and I 286 00:13:47,840 --> 00:13:51,600 guess I'm not a student of which one is better. 287 00:13:51,600 --> 00:13:55,400 I think what's amazing to me about O3 is how great it is like 288 00:13:55,400 --> 00:13:58,440 searching and sort of pulling a bunch of stuff together along 289 00:13:58,440 --> 00:14:01,760 with reasoning, I think. And we're going to make a 290 00:14:01,760 --> 00:14:03,560 concerted effort to prove this true. 291 00:14:03,880 --> 00:14:09,040 I think we can get a genuine scoop thanks to ChatGPT, which I 292 00:14:09,040 --> 00:14:11,680 think would be a landmark moment from reporters. 293 00:14:11,680 --> 00:14:15,120 I don't know how much that's sustainable versus like there's 294 00:14:15,120 --> 00:14:18,240 a lot of stuff that's sort of like on the Internet, the search 295 00:14:18,240 --> 00:14:22,040 isn't as good at finding the now ChatGPT, with its ability to 296 00:14:22,080 --> 00:14:25,200 analyze stuff, can sort of put pieces together. 297 00:14:25,440 --> 00:14:26,480 I don't know how. Exactly. 298 00:14:26,480 --> 00:14:28,760 I was going to say scoops were like the, I mean, scoops are 299 00:14:28,760 --> 00:14:32,040 like the the one thing that AI doesn't have on writers, right, 300 00:14:32,080 --> 00:14:32,920 you know? Right, exactly. 301 00:14:32,920 --> 00:14:34,480 If they can do that, then it's AGI. 302 00:14:34,480 --> 00:14:36,240 That's what I'm saying from my perspective. 303 00:14:36,560 --> 00:14:40,200 I actually don't think that they will be able to turn out scoops 304 00:14:40,200 --> 00:14:46,240 because you pay $200 a month for ChatGPT and that's much less 305 00:14:46,240 --> 00:14:48,280 than you're paying me for scoops. 306 00:14:48,640 --> 00:14:52,600 And I'm very concerned that you think $200 a month equals the 307 00:14:52,600 --> 00:14:54,680 scoops that I could be able to give you. 308 00:14:54,680 --> 00:14:56,960 So by that metric, I do not believe. 309 00:14:57,320 --> 00:14:59,520 That they're able to get scoops says otherwise. 310 00:14:59,640 --> 00:15:02,280 You're yeah, right. Just that's true. 311 00:15:02,280 --> 00:15:04,240 But investment? Wise speaking, it's just not 312 00:15:04,240 --> 00:15:06,920 possible because otherwise you would just pay for that and not 313 00:15:06,920 --> 00:15:09,360 me. You know, it's, it's common for 314 00:15:09,400 --> 00:15:13,040 you, you know is what. Yeah, I believe that's not true 315 00:15:13,040 --> 00:15:15,240 and I believe it can't do that. And there's a huge amount of 316 00:15:15,240 --> 00:15:19,280 value for paying people to write scoops, Significantly more, 317 00:15:19,520 --> 00:15:23,280 significantly more than the $200 a month that you were paying 318 00:15:23,280 --> 00:15:24,200 for. For cheek it is. 319 00:15:24,200 --> 00:15:26,840 Partially tongue, tongue in cheek, but like, I mean, I do 320 00:15:26,840 --> 00:15:30,440 think there's a reality that like lovable can code, but like, 321 00:15:30,480 --> 00:15:32,760 I'm not a coder and so I'm not seeing it through. 322 00:15:32,800 --> 00:15:36,400 And there's going to be the same way that Chachi BT can do some 323 00:15:36,400 --> 00:15:38,400 of these things. But if you're not like a 324 00:15:38,560 --> 00:15:41,440 craftsperson, you don't really have the taste for it. 325 00:15:41,440 --> 00:15:43,960 You don't know what to look for. It's sort of like it's hard to 326 00:15:43,960 --> 00:15:46,240 really extract it. And so I think some of the 327 00:15:46,840 --> 00:15:50,080 threats of job replacement in these sort of areas where 328 00:15:50,080 --> 00:15:52,160 there's a lot of taste and you have to know all the sort of 329 00:15:52,160 --> 00:15:56,880 steps, even if it does some of the steps well, it's, it's only 330 00:15:56,880 --> 00:15:59,680 going to be a tool for the craftspeople. 331 00:16:00,560 --> 00:16:02,480 Yeah. I mean, what was interesting 332 00:16:02,480 --> 00:16:05,280 with the stuff that you were, you know, going on about on 333 00:16:05,280 --> 00:16:07,920 Slack, basically? You know, it can find filings 334 00:16:07,920 --> 00:16:10,280 for like venture Performance. So we hadn't found it right, It 335 00:16:10,320 --> 00:16:12,760 can find. It can find unlisted links that 336 00:16:12,760 --> 00:16:16,400 are listed by, you know, like these venture the LP's and the 337 00:16:16,400 --> 00:16:20,160 endowments and, you know, pension funds where they have 338 00:16:20,160 --> 00:16:21,840 these filings already on the web. 339 00:16:21,840 --> 00:16:25,360 But you have to know the URL to find the specific filing 340 00:16:25,360 --> 00:16:27,440 document. So it's not, you know, fully get 341 00:16:27,440 --> 00:16:30,800 guarded, but it's, you know. And it and it analyzed some 342 00:16:30,800 --> 00:16:33,280 filings from Amazon. I don't think we would ever do 343 00:16:33,280 --> 00:16:34,920 this story. It analyzed some findings from 344 00:16:34,960 --> 00:16:39,960 Amazon that broke down the nature of Amazon's investment in 345 00:16:39,960 --> 00:16:43,960 Anthropic that, you know, Reuters or wherever who had 346 00:16:43,960 --> 00:16:46,640 written the story hadn't really gotten the full breakdown. 347 00:16:46,640 --> 00:16:50,040 And so it argued that there were like nuances of how the 348 00:16:50,040 --> 00:16:53,320 investment was structured, that it caught on to that, that 349 00:16:53,320 --> 00:16:55,200 reporters hadn't yet found which. 350 00:16:55,200 --> 00:16:57,560 So it's already doing like incremental scoops. 351 00:16:57,560 --> 00:16:59,840 It's already doing, you know, like actually this is new 352 00:16:59,840 --> 00:17:01,760 information. It's arguing with its editor. 353 00:17:02,440 --> 00:17:05,200 It's the needle, you know? That's honestly the strongest 354 00:17:05,200 --> 00:17:08,200 case for AGI I've seen yet. If it's already like a very 355 00:17:08,200 --> 00:17:12,079 self-conscious in, you know, sensitive reporter that's trying 356 00:17:12,079 --> 00:17:14,119 to say, no, this is new Intel, this is good stuff. 357 00:17:14,119 --> 00:17:15,280 Why? Why do you think this is just 358 00:17:15,280 --> 00:17:18,119 the same thing Reuters did? I've never more deeply related 359 00:17:18,119 --> 00:17:21,040 to an AI. No, I, I seriously though, I, I 360 00:17:21,040 --> 00:17:23,760 think there's an interesting case to make that if this 361 00:17:23,760 --> 00:17:27,119 somehow does get more wrapped up into the reporting process and 362 00:17:27,280 --> 00:17:30,800 the pulling of public documents or semi public documents is 363 00:17:30,800 --> 00:17:34,160 something that like O3 or O3 Pro can do pretty well. 364 00:17:34,160 --> 00:17:37,160 I think it puts more pressure on the aspects of reporting, which 365 00:17:37,160 --> 00:17:39,560 is gathering information from human beings. 366 00:17:39,880 --> 00:17:42,200 And it's not just document diving which is value. 367 00:17:42,480 --> 00:17:44,120 Like I feel like one of those woke people I need to be 368 00:17:44,120 --> 00:17:46,600 snapping to you. I couldn't agree agree more with 369 00:17:46,600 --> 00:17:48,440 what you're saying. Yeah, that's why we're video 370 00:17:48,440 --> 00:17:50,520 too. Exactly. 371 00:17:50,600 --> 00:17:51,960 I mean this is about the. People. 372 00:17:52,080 --> 00:17:53,880 AI is about the people. AI makes us more. 373 00:17:53,880 --> 00:17:58,440 Human there's 2 levels about it. One like people, the the sort of 374 00:17:58,440 --> 00:18:01,520 like what people are like isn't really like on the Internet. 375 00:18:01,520 --> 00:18:03,840 Like maybe you see YouTube videos, but it's like hard to 376 00:18:04,160 --> 00:18:07,200 it's not a big data problem. It's like a small data problem 377 00:18:07,200 --> 00:18:09,960 where you need to, you know, talk to people, interacting with 378 00:18:09,960 --> 00:18:11,480 people. So getting the story of what 379 00:18:11,480 --> 00:18:14,040 people like and how organizations are working there 380 00:18:14,040 --> 00:18:16,880 isn't like the data for the AI to consume. 381 00:18:17,080 --> 00:18:19,280 And then obviously humans are just better at understanding 382 00:18:19,280 --> 00:18:22,120 like what's interesting about a person and like telling a 383 00:18:22,560 --> 00:18:24,320 personal story. So I totally agree with you. 384 00:18:24,320 --> 00:18:28,800 Like AI will be better at solving data stories and stuff 385 00:18:28,800 --> 00:18:31,240 where it's like pulling together what's what's happening with 386 00:18:31,240 --> 00:18:34,240 like money and numbers. And reporters are going to have 387 00:18:34,240 --> 00:18:36,440 to focus on people like more than ever. 388 00:18:36,600 --> 00:18:40,200 It's interesting to see how this tech will be incorporated into 389 00:18:40,200 --> 00:18:43,120 newsrooms because there's a desperate move by all top 390 00:18:43,120 --> 00:18:44,640 people. And I know this from my time at 391 00:18:44,640 --> 00:18:48,840 the Wall Street Journal to force down the newsroom's throat AI 392 00:18:48,840 --> 00:18:51,200 and into the process because it makes you seem futuristic in 393 00:18:51,200 --> 00:18:53,560 advance. And you can pitch to, you know, 394 00:18:53,560 --> 00:18:56,840 your non journalistic business execs that were embracing AI. 395 00:18:57,320 --> 00:19:00,880 And so far, that's mostly meant using like AI generated talking 396 00:19:00,880 --> 00:19:04,400 points or like bullet points at the tops of stories, which at 397 00:19:04,400 --> 00:19:07,680 best maybe removes one person whose job was to write those 398 00:19:07,680 --> 00:19:10,000 things. But it's not AI. 399 00:19:10,000 --> 00:19:12,640 It's not actually disruptive in the in the flow of news or 400 00:19:12,640 --> 00:19:14,800 reporting. So maybe what you were messing 401 00:19:14,800 --> 00:19:17,520 around with Eric will be closer to it and that can actually 402 00:19:17,520 --> 00:19:19,120 change things. But aside from that, it's been 403 00:19:19,120 --> 00:19:21,840 pretty, I think. It's a good sounding board, good 404 00:19:21,880 --> 00:19:23,520 proofreader. I mean, we still pay a 405 00:19:23,520 --> 00:19:25,160 proofreader, but you know, it's not. 406 00:19:25,280 --> 00:19:26,520 It's going to be more and more helpful. 407 00:19:26,520 --> 00:19:30,040 But I agree that the heavy-handed attempts to to use 408 00:19:30,040 --> 00:19:34,720 it can be heavy-handed. Do you think O3 and similar 409 00:19:34,720 --> 00:19:41,040 models represent AGI as Tyler Cowen seems to say, or as crank 410 00:19:41,040 --> 00:19:44,120 Gary Marcus? Please, this is nowhere close. 411 00:19:44,120 --> 00:19:46,240 Can't even do math with decimal points. 412 00:19:46,240 --> 00:19:49,320 It's idiotic and there's a propaganda machine and. 413 00:19:49,800 --> 00:19:53,240 I, I don't know, man, I don't know that the bar is always 414 00:19:53,240 --> 00:19:54,760 moving. I don't really know what it 415 00:19:54,760 --> 00:19:56,280 means. I mean, there's like financial 416 00:19:56,280 --> 00:19:58,120 definitions for it. There's philosophical 417 00:19:58,120 --> 00:20:00,840 definitions. I have a feeling that like this 418 00:20:01,280 --> 00:20:04,640 back reverse engineering of AGI and say like we've actually had 419 00:20:04,640 --> 00:20:07,160 it for the last couple of years of the last year or so because 420 00:20:07,160 --> 00:20:10,960 of reasoning models that's weak. I'm sorry, Like I, I want a lot 421 00:20:10,960 --> 00:20:13,560 more than that. Like AGI to me needs to be like 422 00:20:13,560 --> 00:20:17,160 in SIM City 2000 when like a headline in the newspaper said, 423 00:20:17,160 --> 00:20:20,080 like nuclear energy exists and you're allowed to like build 424 00:20:20,080 --> 00:20:24,200 nuclear plants and and and. I'm most sympathetic to that, 425 00:20:24,200 --> 00:20:30,600 that it hasn't produced any real innovation or new ideas, and 426 00:20:30,600 --> 00:20:34,480 that is some barrier. But yeah, is does that mean that 427 00:20:34,480 --> 00:20:36,800 the threshold for AGI has to be novelty? 428 00:20:36,800 --> 00:20:41,680 Because I feel like while I'm as a creative person would love to 429 00:20:41,680 --> 00:20:44,200 say yes, that's the threshold and therefore you can't do what 430 00:20:44,200 --> 00:20:47,600 I do, There are plenty of things that it can do. 431 00:20:47,600 --> 00:20:50,560 And if your definition of AGI is to replace someone in the 432 00:20:50,560 --> 00:20:54,120 workforce at a competent level, it's getting closer and closer 433 00:20:54,120 --> 00:20:56,320 to being able to do that to the point where like this could be 434 00:20:56,320 --> 00:20:58,640 the AGI moment. Like I think what you were 435 00:20:58,640 --> 00:21:01,280 working with Eric just for it pulling up filings and things 436 00:21:01,280 --> 00:21:02,800 like that. Like it can do stuff that 437 00:21:02,800 --> 00:21:05,120 analysts can do just hands down already. 438 00:21:05,360 --> 00:21:08,320 I, I think it's safe to say I know a lot about venture capital 439 00:21:08,400 --> 00:21:11,080 and I think open AI knows more than me. 440 00:21:11,080 --> 00:21:14,800 You know, it's just like it can, it can pull all the world's 441 00:21:14,800 --> 00:21:16,720 information and sort of synthesize it. 442 00:21:16,720 --> 00:21:19,720 It makes an occasional error, but so do I. 443 00:21:19,720 --> 00:21:25,480 Like I think even now in areas where I'm an expert, it has more 444 00:21:25,480 --> 00:21:29,520 context and information than me. I mean, I, I, you know, I, I'll 445 00:21:29,520 --> 00:21:31,200 be honest, I've known Gary for a while. 446 00:21:31,200 --> 00:21:32,880 I've probably muted him at this point. 447 00:21:32,880 --> 00:21:35,440 And that's more just out of like respect for him as a person 448 00:21:35,440 --> 00:21:39,720 because I like him, but he's so fucking annoying on X and I'm 449 00:21:39,720 --> 00:21:42,680 willing to keep that in the pod like he just is, even though I 450 00:21:42,680 --> 00:21:45,960 think he's a really smart guy. But I've seen some interesting 451 00:21:45,960 --> 00:21:47,800 stuff on there. Just pointing out like it's 452 00:21:47,800 --> 00:21:51,360 inability to do what would be basic logic that a human, you 453 00:21:51,400 --> 00:21:53,560 know, average and even maybe slightly below average 454 00:21:53,560 --> 00:21:55,960 intelligence could do. There was this one picture, I 455 00:21:55,960 --> 00:21:59,160 don't know if you guys saw it, of like these curvy lines 456 00:21:59,200 --> 00:22:01,360 pointing from a name to like a stick figure. 457 00:22:01,800 --> 00:22:05,280 And it was saying like it, you know, it put the image into O3 458 00:22:05,280 --> 00:22:08,840 and asked it to identify the person with the color of the 459 00:22:08,840 --> 00:22:10,640 stick figure that it was pointing to. 460 00:22:10,640 --> 00:22:12,320 And it got it wrong, completely wrong. 461 00:22:12,720 --> 00:22:16,040 And that's something that probably 1/3 grader could do. 462 00:22:16,720 --> 00:22:20,240 And I think those sorts of mistakes are meaningful. 463 00:22:20,560 --> 00:22:23,760 Like sure, it has research tool capabilities. 464 00:22:23,840 --> 00:22:27,560 And I think, you know, this gets into like the copilot versus 465 00:22:27,600 --> 00:22:30,400 agent debate, like is it something that works as an 466 00:22:30,400 --> 00:22:32,280 assistant or is it like autonomous? 467 00:22:32,640 --> 00:22:36,680 And I think those mistakes, until they're fixed on a regular 468 00:22:36,680 --> 00:22:39,320 enough basis, stopped me from getting too excited about like 469 00:22:39,320 --> 00:22:41,920 human level intelligence. It just doesn't meet. 470 00:22:42,040 --> 00:22:45,320 It just doesn't meet that. All right, we've looked at how 471 00:22:45,320 --> 00:22:47,520 foundation models have been doing really well. 472 00:22:48,080 --> 00:22:53,240 Madeline, your story for Friday looks at RVC is reacting to 473 00:22:53,360 --> 00:22:55,680 market force. Well, they love to say that they 474 00:22:55,760 --> 00:22:57,760 do not react to the markets, right, Eric. 475 00:22:57,760 --> 00:23:00,400 They love to say that they're, you know, thinking of the future 476 00:23:00,400 --> 00:23:03,920 and that the markets now don't really interplay with what 477 00:23:03,920 --> 00:23:06,000 they're doing. True, true or false? 478 00:23:06,080 --> 00:23:07,720 Or what do you? What have you found? 479 00:23:08,000 --> 00:23:11,120 Charles Hudson, a precursor, told me best and said the 480 00:23:11,120 --> 00:23:15,040 tailwind of AI is overcoming the headwinds of the market at the 481 00:23:15,040 --> 00:23:16,440 late stage. They're delayed IP OS. 482 00:23:16,440 --> 00:23:18,640 We cannot disregard that. Obviously that's happening. 483 00:23:19,280 --> 00:23:22,200 I would not say that late stage is 100% affected. 484 00:23:22,200 --> 00:23:23,840 I'd say that there's more of a pause there. 485 00:23:24,080 --> 00:23:27,520 But like at early stage deals, deals are still flowing. 486 00:23:27,560 --> 00:23:30,680 Companies are getting backed. I most of the investors I talked 487 00:23:30,720 --> 00:23:34,280 to had not, you know, increased the pace of deals that they plan 488 00:23:34,280 --> 00:23:37,760 to do over the next few months and by the end of the year, but 489 00:23:37,760 --> 00:23:39,720 most of them also had not slowed down. 490 00:23:39,800 --> 00:23:42,400 The money is still flowing. It's just, of course, you know, 491 00:23:42,480 --> 00:23:44,600 the classic AI versus everybody else story. 492 00:23:44,720 --> 00:23:48,640 Do you think they're in denial about how bad the economy could 493 00:23:48,640 --> 00:23:50,400 get? I mean, Ryan Peterson, who 494 00:23:50,400 --> 00:23:53,360 everybody in Silicon Valley trust we had on the podcast, is 495 00:23:53,360 --> 00:23:56,400 like, if these tariffs don't get reversed, like tons of 496 00:23:56,400 --> 00:23:59,680 businesses are going to die. Like we're already seeing, you 497 00:23:59,680 --> 00:24:02,600 know, companies sell to their factory in China. 498 00:24:02,600 --> 00:24:04,560 Like, obviously those aren't tech companies. 499 00:24:04,560 --> 00:24:07,560 Those are like, you know, T-shirt manufacturers probably, 500 00:24:07,560 --> 00:24:09,480 but that could have a huge impact. 501 00:24:09,480 --> 00:24:11,880 Or like, how much do you think they're in denial about how bad 502 00:24:12,200 --> 00:24:15,880 the economy could get? I mean, I think that the early 503 00:24:15,880 --> 00:24:18,560 stage investors, it's not as much as they're in denial, they 504 00:24:18,560 --> 00:24:20,400 just like are looking at a 10 year horizon. 505 00:24:20,400 --> 00:24:23,480 So like in their mind, maybe the economy's fixed in 10 years, who 506 00:24:23,480 --> 00:24:24,800 cares? This is going to be the next 507 00:24:24,880 --> 00:24:27,240 generational wave of technology and we got to get in it right 508 00:24:27,240 --> 00:24:29,360 now. I'd say more at the Series B 509 00:24:29,360 --> 00:24:32,240 plus level, there's a little more doom and gloom. 510 00:24:32,240 --> 00:24:35,640 It's not entirely rosy. I've heard a couple rumors of 511 00:24:35,640 --> 00:24:39,680 term sheets getting pulled for companies in logistics, chip 512 00:24:39,680 --> 00:24:42,600 making tools, anything like kind of that touches things that'll 513 00:24:42,600 --> 00:24:45,040 be heavily tariffed that are raising right now. 514 00:24:45,440 --> 00:24:47,280 I don't think they're it's not that they're not going to be 515 00:24:47,280 --> 00:24:48,760 able to raise. It's just looking a little 516 00:24:48,760 --> 00:24:51,400 tougher. Anything in retail not looking 517 00:24:51,400 --> 00:24:53,160 so great. So yes, there are carve outs 518 00:24:53,160 --> 00:24:55,480 where there are like things that are not doing well, but there's 519 00:24:55,480 --> 00:24:58,760 also like the Consumer Packaged Goods companies like the D2C 520 00:24:58,760 --> 00:25:01,320 marketplaces that were like really big in the boom times. 521 00:25:01,480 --> 00:25:04,280 They haven't been doing super well honestly for a while. 522 00:25:04,280 --> 00:25:07,240 And this is just kind of like nail in the coffin for them, 523 00:25:07,240 --> 00:25:09,360 unfortunately. You know, if you're sticking to 524 00:25:09,360 --> 00:25:14,560 like VC, like software company, AI engineering, AI application, 525 00:25:14,680 --> 00:25:17,240 those deals are fine. That has not slowed down at all. 526 00:25:17,240 --> 00:25:20,160 No one, especially in early stage VC is like worried about 527 00:25:20,160 --> 00:25:23,480 the markets right now other than just like when you get to that 528 00:25:23,480 --> 00:25:26,760 sort of mid stage is when people start to like slow down a bit. 529 00:25:26,960 --> 00:25:30,280 If you had to do business, tech is where you'd want to do it. 530 00:25:30,320 --> 00:25:32,600 You know, the least physical goods possible. 531 00:25:32,600 --> 00:25:35,840 I mean, obviously what Tom you, you've touched on the fact that 532 00:25:36,280 --> 00:25:39,200 software is not just ones and zeros. 533 00:25:39,200 --> 00:25:42,760 There is like a hardware component that influences the 534 00:25:42,760 --> 00:25:45,440 cost structure. What's your view on how much 535 00:25:45,720 --> 00:25:48,560 worse things are going to get for the tech industry? 536 00:25:48,720 --> 00:25:51,080 I mean, I feel pretty good right now. 537 00:25:51,080 --> 00:25:55,200 Patriots are in control. You know, we've got the All In 538 00:25:55,240 --> 00:25:57,600 podcast, people mostly running policy. 539 00:25:58,520 --> 00:26:00,720 It's kind of hard for me to feel more confident than that. 540 00:26:00,720 --> 00:26:02,600 It seems like everything's gone according to plan. 541 00:26:02,880 --> 00:26:04,920 So I don't know where any of this doom and gloom is coming 542 00:26:04,920 --> 00:26:06,200 from. I mean, look, if you want to get 543 00:26:06,200 --> 00:26:08,000 things done, you've got to do things. 544 00:26:08,000 --> 00:26:10,680 It doesn't matter what they are. You just got to do things. 545 00:26:11,000 --> 00:26:13,200 And I think we're in the era of doing things. 546 00:26:13,480 --> 00:26:16,240 And I couldn't be more impressed with the administration's do 547 00:26:16,240 --> 00:26:19,440 things, whether the thing that they do undoes the previous 548 00:26:19,440 --> 00:26:22,800 thing doesn't matter as long as it was a thing that was done. 549 00:26:22,800 --> 00:26:25,280 And I also want to say I'm very impressed with the number of 550 00:26:25,280 --> 00:26:28,520 Twitter ads that I'm seeing that then have community notes saying 551 00:26:28,760 --> 00:26:31,640 this is a scam propagated by a drop shipping company. 552 00:26:32,640 --> 00:26:34,240 Well, they're. All I think drop shipping is 553 00:26:34,240 --> 00:26:36,240 getting killed. You're literally seeing tweets 554 00:26:36,240 --> 00:26:39,680 of like Donald Trump. I I voted for you, I love you, I 555 00:26:39,680 --> 00:26:43,000 worship you, you can do no wrong, but you're killing my 556 00:26:43,000 --> 00:26:44,320 business. Please. 557 00:26:44,800 --> 00:26:47,800 Yeah, yeah, Trump's completely turning his back on the whole 558 00:26:47,800 --> 00:26:50,480 drop shipping community I think was a huge mistake. 559 00:26:51,040 --> 00:26:53,360 Granted, these people don't seem to have money or it's not 560 00:26:53,360 --> 00:26:55,200 denominated in dollars, it seems. 561 00:26:55,320 --> 00:26:59,360 But I'm, I really think like Trump versus the drop shippers. 562 00:26:59,360 --> 00:27:03,080 Like I don't want to make any predictions, but like I I, I, I 563 00:27:03,120 --> 00:27:05,560 make sure I keep a close eye on the drop shippers if I would 564 00:27:05,600 --> 00:27:07,000 farmers. Are going to get a bailout. 565 00:27:07,000 --> 00:27:09,480 Maybe they'll do a targeted bailout on the drop shippers. 566 00:27:09,720 --> 00:27:12,160 If there's a if there's a drop shipping bailout, that would be 567 00:27:12,360 --> 00:27:15,320 incredible. Drop shipping Bailout. 568 00:27:15,520 --> 00:27:17,760 I we're we're cooked. It's over. 569 00:27:17,840 --> 00:27:20,680 I'm sorry. I want Liberation Day 2 is going 570 00:27:20,680 --> 00:27:23,120 to be a huge press conference where like everybody that's 571 00:27:23,120 --> 00:27:28,080 selling, you know, stuff that was a drop ship from Vietnam is 572 00:27:28,840 --> 00:27:31,120 that that is the end state of America that I've been very 573 00:27:31,120 --> 00:27:32,200 excited about. So. 574 00:27:32,680 --> 00:27:34,840 I. That's as much as I can offer on 575 00:27:34,840 --> 00:27:37,320 this one. Well, with that last little dash 576 00:27:37,320 --> 00:27:40,320 of Tom nihilism, I think we pretty much covered the whole 577 00:27:40,320 --> 00:27:44,160 gambit of news this week. We got, you know, the foundation 578 00:27:44,160 --> 00:27:46,960 models crushing it in terms of business. 579 00:27:46,960 --> 00:27:51,520 We've got VCs excited, markets be damned, except maybe a little 580 00:27:51,520 --> 00:27:53,760 bit worried, but we won't talk about it now because we're still 581 00:27:53,760 --> 00:27:55,680 doing deals. Eric, you had a nice chat with 582 00:27:55,880 --> 00:27:58,520 Kyle Harrison about VC Polyamory. 583 00:27:58,640 --> 00:28:01,200 How did that go? Yeah, he he did great thought 584 00:28:01,200 --> 00:28:04,080 leadership. He made it a viral image on 585 00:28:04,080 --> 00:28:06,160 Twitter. The ultimate influencer. 586 00:28:06,240 --> 00:28:08,160 That's like a research report in venture anyway. 587 00:28:08,360 --> 00:28:12,880 So Kyle Harrison at Contrary friend of the pod, came on to 588 00:28:12,880 --> 00:28:16,600 talk about venture capitalists investing in many, many 589 00:28:16,600 --> 00:28:18,600 foundation model companies in his chart of all the 590 00:28:18,600 --> 00:28:21,840 incestuousness in venture and foundation model companies. 591 00:28:21,840 --> 00:28:25,760 So it was a fun coda to Tom's reporting on the state of 592 00:28:25,760 --> 00:28:27,520 models. Kyle and I talked about how 593 00:28:27,760 --> 00:28:30,480 venture capitalists are playing all the money flowing into 594 00:28:30,800 --> 00:28:32,760 foundation model companies. Give it a listen. 595 00:28:33,000 --> 00:28:35,920 Hey, it's Eric Newcomer. I'm here with Kyle Harrison of 596 00:28:35,920 --> 00:28:42,040 Contrary, former investor at Index and CO2 Thanks for joining 597 00:28:42,040 --> 00:28:43,120 me. Thanks for having me. 598 00:28:43,360 --> 00:28:48,160 We we are here to talk about one specific thing, infidelity among 599 00:28:48,160 --> 00:28:52,160 venture capitalists. Where you, you had a viral image 600 00:28:52,160 --> 00:28:56,960 in tech world that shows all the different investments that VCs 601 00:28:56,960 --> 00:28:59,720 are making into foundation model companies. 602 00:28:59,720 --> 00:29:02,440 And you know, a lot of the VCs are making investments into 603 00:29:02,440 --> 00:29:04,440 competitors. So fun topic. 604 00:29:04,440 --> 00:29:07,960 Do you want to just start off like the impetus for making the 605 00:29:07,960 --> 00:29:11,400 graphic and and then sort of walk us through what you found? 606 00:29:11,760 --> 00:29:15,280 It was kind of a vibe that I had been noticing over the course of 607 00:29:15,280 --> 00:29:18,360 several months where I would see these fundraising announcements 608 00:29:18,360 --> 00:29:20,920 and it would be a big firm talking about, you know, I think 609 00:29:20,920 --> 00:29:23,440 it was XAI. I was the one that really 610 00:29:23,440 --> 00:29:25,680 triggered me where I saw people that were like, you know, we're 611 00:29:25,720 --> 00:29:28,040 so excited to back you on again, you generational company 612 00:29:28,040 --> 00:29:29,480 building all that. And it was all this like stuff 613 00:29:29,480 --> 00:29:30,760 about how great the business was. 614 00:29:31,040 --> 00:29:34,000 And I was like, weren't wait, didn't these people invest in 615 00:29:34,000 --> 00:29:35,320 opening eye or in traffic or whatever? 616 00:29:35,320 --> 00:29:37,640 And I went and looked and was like, oh, they did like, oh, I 617 00:29:37,640 --> 00:29:41,200 guess people are making a lot of cross multi layered investments 618 00:29:41,200 --> 00:29:43,160 and stuff. And then and then in a non 619 00:29:43,160 --> 00:29:45,960 account that I really like, who's one of the best LP 620 00:29:45,960 --> 00:29:47,920 accounts on here? Is this guy endowment Eddie? 621 00:29:47,920 --> 00:29:50,360 He was like, OK, wait, is backing competitive companies 622 00:29:50,360 --> 00:29:53,000 allowed now in venture? And I was like, let me go look 623 00:29:53,000 --> 00:29:55,000 at the cross section and it's like, yeah, I guess so. 624 00:29:55,160 --> 00:29:58,400 Yes, clearly is the answer. Do you just want to walk us 625 00:29:58,400 --> 00:30:02,640 through it like Andreessen? You have SSI, Ilya's company, 626 00:30:03,320 --> 00:30:07,520 XAI, Open AI and Mira's Thinking Machines. 627 00:30:07,680 --> 00:30:09,640 What are some of the other interesting ones? 628 00:30:10,080 --> 00:30:11,400 Yeah. I mean, I think the biggest 629 00:30:11,400 --> 00:30:13,760 thing was that you've got, and I think even some of these I 630 00:30:13,760 --> 00:30:16,080 missed, like this was not necessarily a purely academic 631 00:30:16,080 --> 00:30:19,560 exercise, but what's crazy. Is you made a screenshot, now 632 00:30:19,640 --> 00:30:21,160 you are the scholar on. This. 633 00:30:21,160 --> 00:30:22,800 That's right. That's the level. 634 00:30:22,800 --> 00:30:25,480 So it works. On the Internet, this is much 635 00:30:25,480 --> 00:30:28,880 more baked than most VC takes with just a screenshot anyway. 636 00:30:28,920 --> 00:30:29,480 Go ahead. Yeah. 637 00:30:29,560 --> 00:30:31,480 No, Harvard sent me a degree in the mail. 638 00:30:32,040 --> 00:30:35,560 Yeah, exactly. I, I feel like the biggest thing 639 00:30:35,560 --> 00:30:38,440 is like, you can kind of clearly see, and I think somebody 640 00:30:38,440 --> 00:30:41,240 actually retweeted this and pointed this out, that it's like 641 00:30:41,240 --> 00:30:44,480 a very specific strategy where like the one who's most active 642 00:30:44,480 --> 00:30:46,320 is Andreessen, which is consistent. 643 00:30:46,320 --> 00:30:48,640 Like Andreessen is super broad and investing in a lot of 644 00:30:48,640 --> 00:30:50,800 different. Martine Casado, who I love, set 645 00:30:50,800 --> 00:30:53,480 on stage at Cerebral Valley when I was like, how do you know 646 00:30:53,480 --> 00:30:56,400 which segment of AI to invest in? 647 00:30:56,760 --> 00:30:57,960 He was just like, why not all of them? 648 00:30:59,040 --> 00:31:01,160 They're they're like openly going for coverage at this 649 00:31:01,160 --> 00:31:03,400 point. And apparently even in the key 650 00:31:03,400 --> 00:31:06,720 categories with competitors. Yeah, that's Andreessen. 651 00:31:07,040 --> 00:31:08,360 So Andreessen I think is everywhere. 652 00:31:08,360 --> 00:31:13,440 You've got Sequoia who is, is also quite active, a little less 653 00:31:13,440 --> 00:31:15,680 than Andreessen. So a little bit more, you know, 654 00:31:16,040 --> 00:31:19,200 demur or whatever. And then I think you have firms 655 00:31:19,200 --> 00:31:21,840 like. Sequoia just did it to I mean, 656 00:31:21,840 --> 00:31:24,280 Sequoia's interesting because they obviously what Affirm and 657 00:31:24,280 --> 00:31:27,920 Chlorina, they have these famous ones where they've done both. 658 00:31:27,920 --> 00:31:31,400 They try, I think, to be loyal, but they would have more of a 659 00:31:31,400 --> 00:31:35,760 story of like, you know, I don't know, it's getting pretty broad. 660 00:31:35,760 --> 00:31:37,680 They're in a lot of them. I was going to try and do some 661 00:31:37,680 --> 00:31:41,200 spin that it's like, oh, they they're an X in that sort of 662 00:31:41,200 --> 00:31:44,440 relationship with the rule off world, and then they invest in 663 00:31:44,440 --> 00:31:47,480 the leader, which is clearly open AI in my opinion. 664 00:31:47,560 --> 00:31:49,560 I don't know. But yeah, what it was Any 665 00:31:49,560 --> 00:31:52,280 Sequoia analysis? I think cuz it's also, you think 666 00:31:52,280 --> 00:31:54,200 about the story of like, I don't know if you remember this, but 667 00:31:54,200 --> 00:31:57,840 they invested in Phoenix back in the day on the payment side and 668 00:31:57,840 --> 00:31:59,360 Stripe was pretty upset about it. 669 00:31:59,600 --> 00:32:02,600 And so Sequoia was literally like our bad keep the money. 670 00:32:02,600 --> 00:32:05,520 We won't take an equity stake. Like it's just a donation, you 671 00:32:05,520 --> 00:32:08,040 know, right. So they're definitely more 672 00:32:08,440 --> 00:32:10,200 thoughtful about where they're investing. 673 00:32:10,480 --> 00:32:13,480 I think to Sequoia's credit, one of the things that they've been 674 00:32:13,480 --> 00:32:16,960 very deliberate about is that they are more selective in the 675 00:32:16,960 --> 00:32:19,600 earlier stages. And somebody else I think 676 00:32:19,600 --> 00:32:22,440 pointed this out in a retweet where it's like once it gets 677 00:32:22,440 --> 00:32:25,720 into the later stages, you're largely just, you're garnering 678 00:32:25,720 --> 00:32:28,840 exposure to assets. Like it's not as much once these 679 00:32:28,840 --> 00:32:32,320 guys are raising, you know, $10 billion plus rounds or whatever, 680 00:32:32,640 --> 00:32:34,320 it's like you're really just getting exposure. 681 00:32:34,480 --> 00:32:37,520 I think Sequoia is a little bit more protective of their 682 00:32:37,520 --> 00:32:39,200 reputation at the earliest stages. 683 00:32:39,840 --> 00:32:43,680 Lightspeed is another one. I mean, anthropic is going to be 684 00:32:43,680 --> 00:32:46,360 one of their biggest positions overall. 685 00:32:46,360 --> 00:32:48,720 Like because you know, I dug into their returns and that was 686 00:32:48,720 --> 00:32:50,880 like an totally didn't realize that. 687 00:32:51,120 --> 00:32:53,840 And then they're in Mistral NXAI. 688 00:32:54,960 --> 00:32:57,600 I think they're also in one other, somebody else, I can't 689 00:32:57,600 --> 00:33:00,760 remember which one, but somebody pointed out that maybe there is 690 00:33:00,760 --> 00:33:04,160 one of on here that I missed. And because Lightspeed has also 691 00:33:04,160 --> 00:33:06,160 been super active. And I, I would argue that I 692 00:33:06,160 --> 00:33:09,400 think Lightspeed has been playing not necessarily like 693 00:33:09,400 --> 00:33:11,440 catch up, but they're, they're really trying to like 694 00:33:11,440 --> 00:33:14,760 deliberately demonstrate that they want to be in this bucket 695 00:33:14,760 --> 00:33:18,440 of like the Andreessen's GCS of the world, like just be very, 696 00:33:18,440 --> 00:33:21,560 very active. What's funny, where is GCGC and 697 00:33:21,560 --> 00:33:24,240 NEA like? If you're going to do the 698 00:33:24,240 --> 00:33:27,760 Andreessen style blast them all approach, shouldn't you be in 699 00:33:27,760 --> 00:33:29,880 the defining category of our time? 700 00:33:30,480 --> 00:33:31,840 Yeah, it's a good question. And I I will. 701 00:33:31,840 --> 00:33:34,360 I will defer back again to the lack. 702 00:33:34,400 --> 00:33:38,760 Of my they. Might be they might be in five 703 00:33:38,760 --> 00:33:41,160 of these and I just didn't want to look up another crutch based 704 00:33:41,160 --> 00:33:42,480 profile. Look up everyone. 705 00:33:42,480 --> 00:33:46,160 OK, they'll they'll they'll be like, no, we're we're sluts in 706 00:33:46,160 --> 00:33:50,440 this ad. Who? 707 00:33:50,480 --> 00:33:53,760 Who are the serial monogamists here or who who do you think of 708 00:33:53,760 --> 00:33:57,600 the sort of big brand firms has sort of stuck to one bet? 709 00:33:58,160 --> 00:34:01,320 Well, I think that and now it was actually a sort of oversight 710 00:34:01,320 --> 00:34:03,760 on my side like in making this chart because originally it was 711 00:34:03,760 --> 00:34:06,520 supposed to show the cross section, but I kind of just 712 00:34:06,520 --> 00:34:08,600 started at the bottom. I said, OK, open AI, who's in 713 00:34:08,600 --> 00:34:10,480 that boom, boom. And I worked my way around with 714 00:34:10,480 --> 00:34:12,280 the companies adding firms as I went. 715 00:34:12,760 --> 00:34:15,040 And what I realized when I get when I posted it and it kind of 716 00:34:15,040 --> 00:34:18,679 popped off is that I had left founders found on there despite 717 00:34:18,679 --> 00:34:21,440 the fact that they are actually monogamous in this in this 718 00:34:21,440 --> 00:34:24,400 particular strategy. And the other one that I didn't 719 00:34:24,400 --> 00:34:26,800 put on here, but that that called it out is Thrive. 720 00:34:26,800 --> 00:34:28,920 And I think that again, going back to this like this actually 721 00:34:28,920 --> 00:34:33,199 pretty indicative sampling of different firm strategies. 722 00:34:33,800 --> 00:34:37,440 I think that Founders Fund and Thrive are famous for being 723 00:34:37,440 --> 00:34:41,120 really dedicated to running an incredibly concentrated 724 00:34:41,120 --> 00:34:45,159 portfolio and just like doubling down, tripling down into their 725 00:34:45,159 --> 00:34:46,960 winners. And founders fund, despite being 726 00:34:46,960 --> 00:34:51,159 like perceived as like, I don't know, perceptive and early, you 727 00:34:51,159 --> 00:34:54,080 know, Airbnb was like series C. Some of these they're like, what 728 00:34:54,080 --> 00:34:56,679 is the momentum company that's going to define the generation? 729 00:34:56,679 --> 00:35:00,320 How can we like back up the truck into it? 730 00:35:00,320 --> 00:35:03,840 And you know, clearly they, they see open AI here. 731 00:35:03,880 --> 00:35:06,520 Well, and I think that's one of the reasons why they're not 732 00:35:06,520 --> 00:35:10,320 afraid to admit when they did, even if they didn't get there as 733 00:35:10,320 --> 00:35:12,960 early as possible, right? I think that and probably in 734 00:35:12,960 --> 00:35:15,560 that in that vein of monogamous, I think Khosla is another one 735 00:35:15,560 --> 00:35:18,200 where they've been all open AI all day for AS. 736 00:35:18,200 --> 00:35:21,680 Long as possible, like you roll that company in just right. 737 00:35:22,080 --> 00:35:24,640 Jesus, right? Yeah, that would be insane. 738 00:35:24,640 --> 00:35:28,000 I think to diversify away. It's one thing, you know, I 739 00:35:28,000 --> 00:35:30,040 think a consideration for a lot of these is they're getting an 740 00:35:30,040 --> 00:35:31,920 open AI. It's such a high price that it 741 00:35:31,920 --> 00:35:35,280 fits into the growth story where you're talking about where they 742 00:35:35,280 --> 00:35:37,880 were like, not traditional venture investors in open AI. 743 00:35:37,880 --> 00:35:40,840 They're just like we can't miss the company of our generation 744 00:35:41,160 --> 00:35:43,200 and then they're like, well, let's try and do real venture 745 00:35:43,200 --> 00:35:45,840 investments in other model companies if we get a shot at. 746 00:35:46,480 --> 00:35:48,080 Yeah. So I think the and so, but I 747 00:35:48,080 --> 00:35:51,120 think to founders funds credit, even if they're not there at the 748 00:35:51,120 --> 00:35:54,000 earliest days, they're willing to acknowledge that like, hey, 749 00:35:54,000 --> 00:35:56,920 listen, this is clearly broken out as a category definer and 750 00:35:56,920 --> 00:35:59,520 they'll say that too. Like I've heard Trey say this 751 00:35:59,520 --> 00:36:02,240 multiple times where they talk about, listen, like our 752 00:36:02,240 --> 00:36:04,760 perspective is that there's one company and like, if you're 753 00:36:04,760 --> 00:36:06,840 going to invest in space and you didn't invest in SpaceX, you 754 00:36:06,840 --> 00:36:09,480 probably lost money. If you want to invest in defense 755 00:36:09,480 --> 00:36:11,280 and you didn't invest in Anderol, you probably are going 756 00:36:11,280 --> 00:36:13,000 to lose money. Like they, that's their thought 757 00:36:13,000 --> 00:36:14,560 process. And so even if they're getting 758 00:36:14,560 --> 00:36:16,760 there a little late, they're going to back up the truck. 759 00:36:17,360 --> 00:36:19,200 And that's Peter Thiel's like, you want to be a monopoly 760 00:36:19,200 --> 00:36:20,240 business. It's sort of like you don't. 761 00:36:20,240 --> 00:36:21,920 You don't want to be in a company that's creating a 762 00:36:21,920 --> 00:36:24,040 category necessarily, because then you have all these 763 00:36:24,040 --> 00:36:26,320 competitors. You want to be the company in a 764 00:36:26,320 --> 00:36:29,240 category that chokes out all the oxygen for everybody else. 765 00:36:29,680 --> 00:36:31,640 That's right. I, I, I also think that there is 766 00:36:31,640 --> 00:36:33,840 a narrative here, we can talk about this if you want to, that 767 00:36:34,120 --> 00:36:37,040 there is increasing differentiation. 768 00:36:37,040 --> 00:36:40,080 It's very subtle and it's, and it's going to start overlapping 769 00:36:40,080 --> 00:36:44,920 again, but there is some semblance of different companies 770 00:36:44,920 --> 00:36:46,320 trying to prioritize different places. 771 00:36:46,320 --> 00:36:47,560 We can talk about that if you want to, but. 772 00:36:47,720 --> 00:36:50,520 No, I think, yeah, let's let's talk about it, especially in 773 00:36:50,520 --> 00:36:54,360 light of Opening Eye potentially purchasing windsurf, formerly 774 00:36:54,360 --> 00:36:57,760 known as Codium, that that speaks to them, you know, 775 00:36:57,760 --> 00:37:01,680 demolishing everything, but also speaks to the idea that if 776 00:37:01,680 --> 00:37:04,000 you're an investor, it's hard to know where these companies are 777 00:37:04,000 --> 00:37:05,600 going to go, right? They're sort of going to lean 778 00:37:05,600 --> 00:37:08,720 into where they're successful foundation models in particular. 779 00:37:08,720 --> 00:37:11,160 It's like you could become an application, you could be 780 00:37:11,160 --> 00:37:12,960 infrastructure. You know, it's hard, hard to 781 00:37:12,960 --> 00:37:14,480 know. So are you really? 782 00:37:14,600 --> 00:37:18,040 No one is saying VC shouldn't invest in windsurf. 783 00:37:18,160 --> 00:37:21,080 And if windsurf is part of Opening Eyes turf, then it's 784 00:37:21,080 --> 00:37:22,960 sort of like, you know, the floodgates are open. 785 00:37:22,960 --> 00:37:24,040 Or what would you say about that? 786 00:37:24,560 --> 00:37:27,680 So I think that the, it's actually very, it's very, very 787 00:37:27,680 --> 00:37:30,840 similar to what happened with the hyper scalers where you look 788 00:37:30,840 --> 00:37:33,920 at like the Amazons and Googles and Microsofts of the world. 789 00:37:34,280 --> 00:37:37,640 The difference was that they're sort of amalgamation of a bunch 790 00:37:37,640 --> 00:37:41,080 of different faces happened long after they had been founded, 791 00:37:41,080 --> 00:37:44,200 long after they'd gone public. The difference was it's not a, 792 00:37:44,200 --> 00:37:46,880 it's not necessary because even like think about like open AI 793 00:37:46,920 --> 00:37:51,600 launching a social app and, and I had a friend of mine tweet and 794 00:37:51,600 --> 00:37:54,000 say, so are just like our network effects dead, Like 795 00:37:54,000 --> 00:37:55,520 there's nothing mean anything anymore. 796 00:37:55,520 --> 00:37:58,320 All the businessy stuff that we've learned is, is nonsense. 797 00:37:58,760 --> 00:38:01,120 And it's like, if you think about it, it's not that any of 798 00:38:01,120 --> 00:38:03,120 the fundamental principles have changed. 799 00:38:03,320 --> 00:38:06,960 It's that the velocity or like speed of the thing happening is 800 00:38:06,960 --> 00:38:09,360 just so much faster than it's ever been before. 801 00:38:09,640 --> 00:38:11,920 But at the end of the day, like if open AI does launch a social 802 00:38:11,920 --> 00:38:15,560 app and it does turn out to be successful, it's going to be for 803 00:38:15,560 --> 00:38:19,160 the same reasons why when Meta launched threads, they also got 804 00:38:19,160 --> 00:38:21,920 up there pretty quick is because they have a big broad 805 00:38:21,920 --> 00:38:24,720 established installed base they can they can benefit from. 806 00:38:24,960 --> 00:38:28,240 So it's exactly the same dynamics, but they can play into 807 00:38:28,400 --> 00:38:31,520 different strategies. And, and so when I think about 808 00:38:31,520 --> 00:38:35,040 like each company, each each for a model company, I feel like 809 00:38:35,040 --> 00:38:37,960 they are trying to carve out their niche in the universe. 810 00:38:38,240 --> 00:38:41,600 Some are fuzzier than others, but I do think the clearest 811 00:38:41,600 --> 00:38:45,720 distinction is when you look at the projected revenue of a 812 00:38:45,720 --> 00:38:48,920 specifically Anthropic versus Open Eye. 813 00:38:50,240 --> 00:38:53,040 I think it was like 2027 revenue estimates that that somebody had 814 00:38:53,040 --> 00:38:55,640 leaked or put out or whatever. But it was basically like, hey, 815 00:38:55,640 --> 00:39:00,360 by 2027, Anthropic believed that they would have I think like 5X 816 00:39:00,360 --> 00:39:03,440 the API revenue that Open AI was projecting for. 817 00:39:03,440 --> 00:39:05,320 And it wasn't Anthropic crapping on Open AI. 818 00:39:05,320 --> 00:39:07,880 It was Opening Eyes forecast for 27 compared to Anthropic's 819 00:39:07,880 --> 00:39:10,160 forecast. And Anthropic assumes they're 820 00:39:10,160 --> 00:39:13,600 going to have five times as much API revenue than Open AI. 821 00:39:13,960 --> 00:39:17,440 And it becomes really obvious that Open AI, whether they meant 822 00:39:17,440 --> 00:39:23,000 to do this or not, is going to become largely a consumer driven 823 00:39:23,040 --> 00:39:25,680 business. Like ChatGPT is their cash cow 824 00:39:25,720 --> 00:39:28,680 right now. And whereas Anthropic is trying 825 00:39:28,680 --> 00:39:32,960 to squarely plant itself in the API where it's sort of the B to 826 00:39:32,960 --> 00:39:35,200 B world. That's a clear distinction 827 00:39:35,200 --> 00:39:37,360 that's defining a lot of the decisions those companies are 828 00:39:37,360 --> 00:39:39,480 making. You know, I love, love Anthropic 829 00:39:39,480 --> 00:39:42,800 and they're very competitive, but there is the, the sort of 830 00:39:42,800 --> 00:39:47,960 infrastructure API business is like under threat from open 831 00:39:47,960 --> 00:39:51,200 source in a way that having a dominant consumer application 832 00:39:51,840 --> 00:39:54,560 is, is a type of Moat that we're all all familiar with. 833 00:39:54,560 --> 00:39:58,480 It feels like besides open AI, having a lot of users, everybody 834 00:39:58,480 --> 00:40:02,440 else is sort of like we need to run faster than our competition 835 00:40:02,440 --> 00:40:05,320 to have sort of incrementally better stuff. 836 00:40:05,320 --> 00:40:08,840 Obviously over time you build customer relationships, sales, 837 00:40:09,480 --> 00:40:12,880 sort of product differentiation. But it is totally, if I mean 838 00:40:12,880 --> 00:40:17,000 that's absolutely the the bear case on Anthropic is that 839 00:40:17,000 --> 00:40:20,160 progress is just too quick for you to get in and establish any 840 00:40:20,160 --> 00:40:22,240 real meaningful like enterprise install base. 841 00:40:22,920 --> 00:40:26,120 Whereas I think on the consumer side that's different because 842 00:40:26,120 --> 00:40:30,240 when you see these surveys that say something like 6575% of 843 00:40:30,240 --> 00:40:33,800 adults in the US like not in tech, but like period, everyone 844 00:40:34,080 --> 00:40:37,720 has used an AI product. The vast majority of them are 845 00:40:37,720 --> 00:40:39,960 ChatGPT. And I think about this all the 846 00:40:39,960 --> 00:40:43,200 time that like when we look back at the end of 2022 when Chad TPT 847 00:40:43,200 --> 00:40:45,840 came out, What's so interesting is when you talk to the people 848 00:40:45,840 --> 00:40:49,040 who are the sort of experts in this world at that point in 849 00:40:49,040 --> 00:40:52,400 time, that was not a technological breakthrough. 850 00:40:52,400 --> 00:40:54,560 Like it wasn't, that wasn't something that a lot of the 851 00:40:54,560 --> 00:40:57,640 companies who were operating in 2022 couldn't have done. 852 00:40:57,640 --> 00:41:00,400 Like they could have put a chat interface on their model, but 853 00:41:00,400 --> 00:41:03,840 they didn't open AI put out it as like almost an experimental 854 00:41:03,840 --> 00:41:06,080 tool to say, hey, in this neat, you can kind of play around with 855 00:41:06,080 --> 00:41:07,720 it. And it exploded because it was 856 00:41:07,720 --> 00:41:10,600 people's first direct experience with generative text. 857 00:41:10,960 --> 00:41:13,480 Everybody was too afraid of their tails, and only open AI 858 00:41:13,480 --> 00:41:17,800 was willing to unleash AI onto us all in a cavalier way. 859 00:41:17,840 --> 00:41:19,120 Yeah. Yeah, well, I think they even 860 00:41:19,120 --> 00:41:21,120 think it was an accident, right? They like they didn't expect it 861 00:41:21,120 --> 00:41:23,880 to do that either. Right, But I mean, clearly, I 862 00:41:23,880 --> 00:41:26,120 think some of the break of anthropic was, you know, the 863 00:41:26,120 --> 00:41:28,480 roll out of Chachi BT and everything. 864 00:41:29,360 --> 00:41:35,000 I, I do think there's a lot of room for consumer adoption just 865 00:41:35,000 --> 00:41:38,000 because I, I feel like I'm begging people to use open AI or 866 00:41:38,000 --> 00:41:41,040 to use Chachi BT in some case. I'm like, like my family members 867 00:41:41,040 --> 00:41:43,400 are like, it's crazy. I'm like, you know, super 868 00:41:43,400 --> 00:41:47,240 excited about O3 spending a lot of time on it and it it's, you 869 00:41:47,240 --> 00:41:49,800 know, people, these companies clearly need to find ways to 870 00:41:49,800 --> 00:41:52,920 bring the product to people in a way that doesn't take as much 871 00:41:53,200 --> 00:41:54,440 work. And that's going to be a huge 872 00:41:54,440 --> 00:41:56,720 build up going back to venture because that's really the focus 873 00:41:56,720 --> 00:41:59,600 of this conversation. Do you think what we're seeing 874 00:41:59,600 --> 00:42:04,760 with the models is playing out in other categories or do you 875 00:42:04,760 --> 00:42:09,720 think this category is unique where VCs invest all over the 876 00:42:09,720 --> 00:42:14,680 place? So I had one of my much smarter 877 00:42:14,680 --> 00:42:18,520 than me friends kind of articulate this analogy of like 878 00:42:18,520 --> 00:42:22,720 when you think about the.com, there was sort of, you know, the 879 00:42:22,720 --> 00:42:25,360 almost the three layers of what was happening. 880 00:42:25,400 --> 00:42:28,200 And that really drove like the distribution of the Internet. 881 00:42:28,760 --> 00:42:31,000 And you basically have there were folks on the infrastructure 882 00:42:31,000 --> 00:42:34,800 side you're building physical, you know, almost like compute 883 00:42:34,800 --> 00:42:36,800 and stuff like that. Then you have the networking 884 00:42:37,560 --> 00:42:40,320 where it's the Cisco's of the world getting to broaden out and 885 00:42:40,320 --> 00:42:42,480 get access to the underlying tech. 886 00:42:42,840 --> 00:42:44,280 And then you had the applications, right, the 887 00:42:45,040 --> 00:42:48,600 pets.com and stuff like that. The difference there when you 888 00:42:48,600 --> 00:42:51,480 think about like how that is similar to today is that you 889 00:42:51,480 --> 00:42:53,120 should you look at the chips company and chips company once 890 00:42:53,120 --> 00:42:54,680 again are ripping, right. The sort of underlying 891 00:42:54,680 --> 00:42:56,960 infrastructure is still powerful regardless of what's getting 892 00:42:56,960 --> 00:42:58,680 built on top of it. That's very powerful. 893 00:42:58,920 --> 00:43:02,400 And so I think that segment has more problems with, you know, 894 00:43:02,480 --> 00:43:04,960 geopolitical stuff than it does like distribution. 895 00:43:05,560 --> 00:43:08,000 But what's what's really unique about this point in time that 896 00:43:08,000 --> 00:43:10,840 makes it really difficult for venture investors is that in 897 00:43:10,840 --> 00:43:13,160 the.com you had networking companies that were like, we're 898 00:43:13,160 --> 00:43:14,600 just trying to get this out into the world. 899 00:43:14,800 --> 00:43:16,760 And then you had applications coming and saying we're going to 900 00:43:16,760 --> 00:43:19,840 get this package and deliver it to people as a service. 901 00:43:20,440 --> 00:43:23,880 Today you have the networking layer, which is effectively the 902 00:43:23,880 --> 00:43:26,360 model companies also building the application. 903 00:43:26,360 --> 00:43:29,600 So it's effectively as if Cisco was launching pets.com, it would 904 00:43:29,600 --> 00:43:32,040 be really difficult for pets.com to compete with that. 905 00:43:32,200 --> 00:43:34,280 You know, and you know, maybe it falls apart because it wasn't 906 00:43:34,280 --> 00:43:37,040 the underlying thing. But today it is like AI is the 907 00:43:37,040 --> 00:43:39,000 underlying aspect of that product. 908 00:43:39,320 --> 00:43:43,760 And so when you see folks like open AI buying Winsurf like that 909 00:43:43,760 --> 00:43:46,720 is a it's a perfectly logical extension, right? 910 00:43:46,720 --> 00:43:49,000 You see this? To reframe what you're saying a 911 00:43:49,000 --> 00:43:52,720 little bit, you're saying the non monogamy or the polyamory 912 00:43:52,720 --> 00:43:55,720 with foundation models, it is polyamory with everything 913 00:43:55,720 --> 00:43:58,360 because the models sort of are competing with everything that's 914 00:43:58,760 --> 00:44:00,760 investable and interesting right now. 915 00:44:01,080 --> 00:44:02,880 And so it's like, yeah, if you're doing applications, 916 00:44:02,880 --> 00:44:05,440 infrastructure, whatever, inevitably you're going to be a 917 00:44:05,440 --> 00:44:07,600 collision course with your model investment. 918 00:44:08,240 --> 00:44:10,680 That that seems OK. It's just I think it's just the 919 00:44:10,680 --> 00:44:13,480 pure model on model that feels like, Oh my God, like whatever, 920 00:44:14,160 --> 00:44:16,760 you know, how many, how many closed source models do we need? 921 00:44:16,760 --> 00:44:20,440 I guess is a backdrop question. Well, and I guess the question, 922 00:44:20,440 --> 00:44:23,800 I think the, the thing that people are sort of struggling 923 00:44:23,800 --> 00:44:29,720 with is that like the, the not just the progress of individual 924 00:44:29,720 --> 00:44:31,840 companies, but the progress from model to model. 925 00:44:31,840 --> 00:44:35,240 I mean, you look at like people's reactions from Claude 926 00:44:35,240 --> 00:44:38,000 3.5 to 3.7 and it's like that's a miss. 927 00:44:38,000 --> 00:44:41,640 That's a misstep that that like could make or break the game for 928 00:44:41,640 --> 00:44:43,360 a while, right? Like that it can happen that 929 00:44:43,360 --> 00:44:45,040 quickly. Or like, you know, I think 930 00:44:45,040 --> 00:44:48,160 somebody else tweeted this of this like constant like we're so 931 00:44:48,160 --> 00:44:50,600 back. It's so over cycle of like, oh, 932 00:44:50,600 --> 00:44:53,520 like I actually think, you know, 4.5 is not very good. 933 00:44:53,520 --> 00:44:56,520 And then it's like 03 is incredible and like it's so 934 00:44:56,520 --> 00:44:59,720 quick with each model that it's so much harder that like at the 935 00:44:59,720 --> 00:45:02,000 end of the day, yeah, if you're a massive multi billion dollar 936 00:45:02,000 --> 00:45:05,120 firm, you've you've got to go get exposure to everything 937 00:45:05,120 --> 00:45:06,800 because you don't know what's going to come out on top. 938 00:45:06,960 --> 00:45:08,800 This is what we want from capitalism, isn't it? 939 00:45:09,320 --> 00:45:11,800 Everybody's competing, lots of money flowing into an 940 00:45:11,800 --> 00:45:14,360 interesting space. And like, yeah, you got to be 941 00:45:14,360 --> 00:45:17,440 the best of the best if you want to extract value. 942 00:45:17,440 --> 00:45:18,160 I don't know. I'm not. 943 00:45:18,840 --> 00:45:22,600 And it's a category where I'm like happy for the world to burn 944 00:45:22,600 --> 00:45:27,280 some capital, you know, trying to crackitsoidontknow.com 945 00:45:27,880 --> 00:45:30,120 theproblemwith.com was like there was no distribution. 946 00:45:30,560 --> 00:45:32,480 It's like they didn't they you couldn't, there wasn't the 947 00:45:32,480 --> 00:45:36,200 customer base when you set up the Internet now in all the 948 00:45:36,200 --> 00:45:38,720 other bubbles, it's like, well, there's still a distribution to 949 00:45:38,720 --> 00:45:40,120 quickly get a bunch of customers. 950 00:45:40,120 --> 00:45:44,000 I mean you, you invoke.com, are you, are you worried about like 951 00:45:44,000 --> 00:45:48,960 a big bubble? So I my biggest concern is this 952 00:45:48,960 --> 00:45:55,760 idea of like, like hype cycles are built on momentum value 953 00:45:55,760 --> 00:45:58,440 cycles, if you will. Like actual value creation 954 00:45:58,680 --> 00:46:01,760 happens when you're able to push through all of the hype and all 955 00:46:01,760 --> 00:46:04,120 the excitement, all the experimentation and get to the 956 00:46:04,120 --> 00:46:06,400 actual creation of business value. 957 00:46:06,960 --> 00:46:09,360 And if there is underlying business value that's consistent 958 00:46:09,360 --> 00:46:11,800 and replicable and changing the way an organization works, then 959 00:46:11,800 --> 00:46:14,200 you can capture some of that value. 960 00:46:14,920 --> 00:46:19,320 My sense is that the volume of capital and the volume of hype 961 00:46:20,000 --> 00:46:24,160 is not commensurate with the our ability to articulate and 962 00:46:24,160 --> 00:46:27,240 execute on actual business value like that. 963 00:46:27,240 --> 00:46:30,440 Does not those feel wildly? And part of what made that 964 00:46:30,440 --> 00:46:33,400 possible this time is every enterprise in the world was 965 00:46:33,400 --> 00:46:37,160 willing to spend money initially to say, oh, how do we figure out 966 00:46:37,240 --> 00:46:39,240 AI? But that could be a sort of 967 00:46:39,240 --> 00:46:42,560 false signal on how much they stick around when they don't get 968 00:46:42,560 --> 00:46:44,400 the value. And that goes back to your 969 00:46:44,400 --> 00:46:47,840 question about like, you know, we, we the, the biggest thing 970 00:46:47,840 --> 00:46:51,480 that we have to unlock is we have to figure out how do we get 971 00:46:51,480 --> 00:46:53,720 people to use this more consistently in a way that 972 00:46:53,720 --> 00:46:56,120 creates value. That's not just true of like you 973 00:46:56,120 --> 00:46:59,280 and I trying to convince our mom to use ChatGPT or whatever. 974 00:46:59,600 --> 00:47:04,800 That's also true. Yeah, we look too similar. 975 00:47:04,800 --> 00:47:08,920 We can't be saying something like that, but like that that 976 00:47:08,920 --> 00:47:12,080 kind of stuff like it's from the consumer side is also true on 977 00:47:12,080 --> 00:47:14,360 the business side. And I think all the time about 978 00:47:14,360 --> 00:47:17,800 like, yeah, we've seen crazy like Rev scale UPS, right? 979 00:47:18,040 --> 00:47:19,760 But it's like, OK, what are probably the three companies 980 00:47:19,760 --> 00:47:22,720 that have made the most money from this this pop? 981 00:47:23,040 --> 00:47:25,600 And it's like NVIDIA, which is like, OK, demand, demand, people 982 00:47:25,600 --> 00:47:28,960 need it, whatever open AI, which is like innate offering it to 983 00:47:28,960 --> 00:47:30,560 people to play around with and use. 984 00:47:31,080 --> 00:47:34,560 And then like Accenture, who's like making an ungodly amount of 985 00:47:34,560 --> 00:47:37,840 money helping people figure out like, what do what do we do with 986 00:47:37,840 --> 00:47:39,120 this? Like, how does this actually 987 00:47:39,120 --> 00:47:41,520 impact our business? But like outside of that, like 988 00:47:41,520 --> 00:47:44,240 there haven't been a ton of case studies, at least that I've seen 989 00:47:44,640 --> 00:47:48,840 of like, oh, like John Deere is suddenly ramping up. 990 00:47:49,000 --> 00:47:51,360 You hear some stories like I think it was the what the Klarna 991 00:47:51,360 --> 00:47:53,560 story of like. Yeah, Klarna and Shopify, but 992 00:47:53,560 --> 00:47:55,520 some of it feels very aspirational, like they want to 993 00:47:55,520 --> 00:47:58,680 be seen as getting it. Yeah, but it's like. 994 00:47:58,840 --> 00:48:01,360 I want, yeah, totally. Listen, I would love for that to 995 00:48:01,360 --> 00:48:02,720 happen. Like I think that like, yeah, 996 00:48:02,720 --> 00:48:04,720 the fears of like job destruction and stuff like that. 997 00:48:04,720 --> 00:48:08,160 Like I agree that it's the same as like Willy Wonka or what's 998 00:48:08,160 --> 00:48:09,680 his name Charlie and the talking factor. 999 00:48:09,680 --> 00:48:12,040 He's dad getting displaced by a robot and then getting a job 1000 00:48:12,040 --> 00:48:13,200 fixing the robot that took his job. 1001 00:48:13,200 --> 00:48:16,280 Like there's there's lots of opportunities for evolution. 1002 00:48:16,880 --> 00:48:18,920 I'm not afraid of that. I want that to happen. 1003 00:48:19,120 --> 00:48:22,200 The question is just is it capable of doing that? 1004 00:48:22,200 --> 00:48:25,040 And I don't, I don't know that we're there, which could set us 1005 00:48:25,040 --> 00:48:28,480 up for a very painful pop. But like you look back at 1006 00:48:28,480 --> 00:48:32,040 the.com and like Amazon and Google camethroughthe.com, like 1007 00:48:32,040 --> 00:48:34,960 there will be businesses that get built that are still 1008 00:48:34,960 --> 00:48:37,800 generational businesses despite the fact that they will 1009 00:48:37,800 --> 00:48:42,000 participate in short term pain. Sam Allman, Opening eye, did 1010 00:48:42,000 --> 00:48:45,560 reportedly try to scare investors off investing. 1011 00:48:45,600 --> 00:48:47,560 You know, he had that list that was reported. 1012 00:48:48,200 --> 00:48:50,080 You think that was just like a total failure or any 1013 00:48:50,080 --> 00:48:53,240 observations about that? Well, So what was interesting is 1014 00:48:53,240 --> 00:48:56,600 if I remember correctly, one of the biggest sort of stinks that 1015 00:48:56,600 --> 00:49:00,640 was made about that list was Glean was was not people not 1016 00:49:00,640 --> 00:49:02,200 investing in Glean and an open AI. 1017 00:49:02,880 --> 00:49:05,240 And here's the thing, like, so to go back to the question of 1018 00:49:05,240 --> 00:49:10,520 like you're not as worried about the infidelity for between like 1019 00:49:10,520 --> 00:49:13,640 the model and application layer. Like I actually don't think that 1020 00:49:13,640 --> 00:49:17,600 you can distinguish because it, it at the end of the day, it 1021 00:49:17,600 --> 00:49:21,240 increasingly it seems that it is going to be the same thing 1022 00:49:21,240 --> 00:49:24,240 because you're not like the again, like your point, like the 1023 00:49:24,240 --> 00:49:27,520 API revenue, it's really difficult to believe that that's 1024 00:49:27,520 --> 00:49:29,520 long term defensible. Like I thought it was 1025 00:49:29,520 --> 00:49:33,800 fascinating when I dug into all of the Cogen agents right where 1026 00:49:33,800 --> 00:49:36,640 you look at like the cursors and windsurfs and Lovables and bolts 1027 00:49:36,640 --> 00:49:38,240 and replics of the world and stuff like that. 1028 00:49:38,320 --> 00:49:40,440 They're all using Clod. Like they at the time they were 1029 00:49:40,440 --> 00:49:43,080 using clod 3.5. This is a couple months ago, but 1030 00:49:43,120 --> 00:49:45,960 they were all using that model. But it's like tomorrow we can 1031 00:49:45,960 --> 00:49:48,560 all switch over to this one because now Gemini is the better 1032 00:49:48,560 --> 00:49:50,360 code Gen. model or whatever like that. 1033 00:49:50,360 --> 00:49:53,880 Progress means that it's going to be incredibly difficult for 1034 00:49:53,880 --> 00:49:56,520 any of these companies to justify their existence at the 1035 00:49:56,520 --> 00:49:59,920 model layer, which means that, well, where do you make money? 1036 00:50:00,160 --> 00:50:03,440 You go pay $3 billion for windsurf because that's where 1037 00:50:03,440 --> 00:50:05,600 you're going to make money because people want to have a 1038 00:50:05,600 --> 00:50:07,760 thing that they use. The bigger question there then 1039 00:50:07,760 --> 00:50:10,800 for, for windsurf users is like, OK, wait, but am I going to be 1040 00:50:10,800 --> 00:50:14,400 stuck now then on GPT models? And it's like not if they want 1041 00:50:14,400 --> 00:50:16,760 to keep their that installed base like they're going to have 1042 00:50:16,760 --> 00:50:18,520 to let people use different models. 1043 00:50:19,520 --> 00:50:22,000 And on that acquisition, there is people are like, oh, you can 1044 00:50:22,000 --> 00:50:24,960 also train your models on the behavior on windsurf. 1045 00:50:24,960 --> 00:50:27,920 So there's an argument that, you know, you could be open AI, 1046 00:50:28,000 --> 00:50:30,840 leave it, leave customers allowed to use whatever model 1047 00:50:30,840 --> 00:50:33,680 they want, and then learn a lot to then make your model better 1048 00:50:33,680 --> 00:50:36,840 so that eventually they want to use your model again. 1049 00:50:37,200 --> 00:50:38,880 Yeah. But if I think about like, OK, 1050 00:50:38,880 --> 00:50:42,120 if I'm open AI and I want to rule the world and I'm not 1051 00:50:42,200 --> 00:50:44,440 because I do think also like again going back to this, like 1052 00:50:44,440 --> 00:50:45,920 some of these companies will differentiate. 1053 00:50:46,320 --> 00:50:53,160 I think that SSI is going to be sort of go off on this like AI 1054 00:50:53,160 --> 00:50:55,600 safety, like let's save the planet stuff. 1055 00:50:55,600 --> 00:50:56,880 And it's like, let's see how that goes. 1056 00:50:56,880 --> 00:50:59,360 I don't know what that's going to evolve to look like, but it's 1057 00:50:59,360 --> 00:51:02,840 clear that open AI is not, you know, again, maybe the narrative 1058 00:51:02,840 --> 00:51:05,960 and the marketing is making AGI accessible for everybody or 1059 00:51:05,960 --> 00:51:07,240 whatever. It's like, right, They want to 1060 00:51:07,240 --> 00:51:08,880 own the world and they want to make a lot of money and they're 1061 00:51:08,880 --> 00:51:10,640 trying to be a for profit so that they can raise more money 1062 00:51:10,640 --> 00:51:12,840 to make more money. Like they're clearly trying to 1063 00:51:12,840 --> 00:51:15,160 own the world. It's like, OK, if I'm open AI 1064 00:51:15,160 --> 00:51:19,600 and I need to justify a hundreds of billions of dollar valuation, 1065 00:51:20,040 --> 00:51:23,240 how am I going to do that? It's not going to be, I promise 1066 00:51:23,240 --> 00:51:26,320 I'll keep making the next best model and like a constant, you 1067 00:51:26,320 --> 00:51:29,920 know, a hamster wheel, it's going to be, I'm going to be the 1068 00:51:29,920 --> 00:51:32,120 thing for your workplace and I'm going to be the thing for your 1069 00:51:32,120 --> 00:51:35,200 planning and your personal life and your AI agent and whatever. 1070 00:51:35,920 --> 00:51:39,000 I, I was going to say initially, open AI is a uniquely 1071 00:51:39,000 --> 00:51:42,880 problematic juggernaut from an investor perspective because it 1072 00:51:42,880 --> 00:51:47,640 is a non profit and you know, because it had the ousted CEO 1073 00:51:47,640 --> 00:51:49,640 and everything. It's like, I can see why you'd 1074 00:51:49,640 --> 00:51:52,960 want to sort of hedge your bets that maybe it's sort of like a 1075 00:51:52,960 --> 00:51:56,600 shakier dominant player than in past generations, but then you 1076 00:51:56,600 --> 00:52:00,280 reflect and you're like, wait, Uber add the Travis stuff like 1077 00:52:00,560 --> 00:52:02,680 Facebook. Mark Zuckerberg was like firing 1078 00:52:02,680 --> 00:52:05,840 his whole executive team and had to like pivot to mobile in some 1079 00:52:05,840 --> 00:52:09,320 ways, like the juggernaut says company was never like so 1080 00:52:09,320 --> 00:52:13,480 clearly anointed as we we write the story, but you know, going 1081 00:52:13,480 --> 00:52:16,920 through it, you're like, man, opening eye could totally failed 1082 00:52:16,920 --> 00:52:20,000 to become a for profit and then I better have an investment 1083 00:52:20,000 --> 00:52:23,640 somewhere else. I wanted to ask and I guess let 1084 00:52:23,640 --> 00:52:26,040 this be sort of the last question. 1085 00:52:27,240 --> 00:52:31,240 What without talking your own book, what should a founder do 1086 00:52:31,320 --> 00:52:33,760 based on this information or like, what do you think? 1087 00:52:34,120 --> 00:52:36,840 Like if everybody's investing in everybody, like to I need to go 1088 00:52:36,840 --> 00:52:40,960 to a CO2 and thrive or it's like I need to just accept it or 1089 00:52:41,040 --> 00:52:42,440 what? What do you What advice would 1090 00:52:42,440 --> 00:52:46,120 you give to a founder faced with this polyamory? 1091 00:52:47,400 --> 00:52:52,160 So increasingly, and I don't know that I'd go so far as to 1092 00:52:52,160 --> 00:52:55,600 say that polyamory is rampant in every category forever, at every 1093 00:52:55,600 --> 00:52:58,080 stage for every firm. Like everybody's, everybody's 1094 00:52:58,080 --> 00:53:00,600 different. But what we're seeing here is a 1095 00:53:00,600 --> 00:53:02,320 symptom of something that you and I have talked about a lot 1096 00:53:02,320 --> 00:53:06,160 before of this like sort of capital agglomerator, like rise 1097 00:53:06,160 --> 00:53:08,880 of the, you know, the black stones of our industry, right? 1098 00:53:08,880 --> 00:53:10,640 Like that's what's happening. This is a Subs. 1099 00:53:10,640 --> 00:53:14,000 This is a. For lack of a better term. 1100 00:53:14,400 --> 00:53:16,960 Yeah, like this is a sub system of that. 1101 00:53:17,160 --> 00:53:21,320 And I think that in in thinking about that because it is true 1102 00:53:21,320 --> 00:53:22,720 that they're going to do more, there's going to be more 1103 00:53:22,720 --> 00:53:24,320 applications that they're investing in that may be 1104 00:53:24,320 --> 00:53:26,080 competitive and whatever because they're just trying to play 1105 00:53:26,080 --> 00:53:28,600 capital. What founders need to understand 1106 00:53:28,600 --> 00:53:31,560 and what I've spent three years writing about on my on my sub 1107 00:53:31,560 --> 00:53:35,280 sack, is that increasingly venture firms, at least the 1108 00:53:35,280 --> 00:53:38,640 business model venture firm has increasingly become more focused 1109 00:53:38,960 --> 00:53:42,080 on asset management. That's not the marketing. 1110 00:53:42,080 --> 00:53:45,080 They still market themselves like they are the scrappy, you 1111 00:53:45,080 --> 00:53:48,200 know, John Doors of the world from the early 2000s or whatever 1112 00:53:48,680 --> 00:53:49,920 and the benchmarks and stuff like that. 1113 00:53:49,920 --> 00:53:51,200 But that's, that's the marketing. 1114 00:53:51,520 --> 00:53:54,640 The business model of most of these firms has become asset 1115 00:53:54,640 --> 00:53:56,880 management. And if they're going to build 1116 00:53:56,880 --> 00:54:01,520 themselves as asset managers, you as a founder need to treat 1117 00:54:01,520 --> 00:54:04,560 them as asset managers. They are not your, even if 1118 00:54:04,560 --> 00:54:06,280 they're on your board, they are not. 1119 00:54:06,280 --> 00:54:07,920 Just because they're on your board does not mean that they're 1120 00:54:07,920 --> 00:54:09,720 your buddy. It does not mean that they're 1121 00:54:09,720 --> 00:54:11,280 your therapist or your sounding board. 1122 00:54:11,520 --> 00:54:14,120 There are firms and there are investors that you can go get 1123 00:54:14,120 --> 00:54:17,560 who will be your confidante and your friend and support and 1124 00:54:17,560 --> 00:54:19,160 whatever. And their firms are set up more 1125 00:54:19,160 --> 00:54:23,640 aligned with your outcomes. These larger firms are not that. 1126 00:54:23,960 --> 00:54:28,120 And so people need to take a, like a crash course in how 1127 00:54:28,120 --> 00:54:31,880 public companies manage the asset managers that buy their 1128 00:54:31,880 --> 00:54:34,440 stock, because that's the way you need to treat some of these 1129 00:54:34,440 --> 00:54:36,600 firms is like that. And that goes to like things 1130 00:54:36,600 --> 00:54:39,800 like information rights and strategy and, and how and who 1131 00:54:39,800 --> 00:54:42,760 you let into what round and what percentage of ownership and how 1132 00:54:42,760 --> 00:54:46,160 much they control of your parata and your, you know, majority 1133 00:54:46,160 --> 00:54:47,800 voting rights and stuff like that. 1134 00:54:47,800 --> 00:54:51,240 Like you need to manage them as the asset managers that they are 1135 00:54:51,400 --> 00:54:53,640 and be very, very careful. It's not bad. 1136 00:54:53,640 --> 00:54:56,920 It's a good weapon that you can use in a in a in a big fight, 1137 00:54:57,320 --> 00:54:58,840 but you have to be really careful. 1138 00:55:00,440 --> 00:55:03,360 Kyle Harrison Contrary, thanks so much for joining us. 1139 00:55:03,720 --> 00:55:04,360 Thanks for having me.