El “momento Sputnik” de China: SeaDance 2.0 y la IA que amenaza con destruir Hollywood (YouTube Video Transcript)

(00:00:00) Your YouTube transcript will appear here (00:00:06) managing partner of Fluent Ventures. (00:00:09) Good. Good evening Alexander. (00:00:11) >> Hello. Pleasure to be with you. (00:00:13) >> It's a pleasure to to have you here at (00:00:16) the office television with us. Um I (00:00:18) would like to ask you Alexander uh (00:00:19) recent advances in AI tools have (00:00:22) triggered a sharp style of of across (00:00:25) software and data stocks uh from an (00:00:28) investment perspective. Alexander, is (00:00:31) the market reacting to a shortterm fear (00:00:34) or to a genuine structural shift in the (00:00:37) shopware business model? (00:00:40) >> It's been quite a week uh with a sea of (00:00:42) red. Um look, I think it's both uh in (00:00:44) the in in this sense. Um AI is a massive (00:00:49) platform shift that offers us the (00:00:51) opportunity to reinvent how many core (00:00:54) software and technology services and (00:00:56) frankly a lot of human services are (00:00:58) redone. And so, yes, it is a massive (00:01:00) long-term shift. And also, um, I think (00:01:03) it is a big short-term reaction, too. (00:01:06) Um, my sense is that the baby is being (00:01:08) thrown out with the bathwater. I think (00:01:09) there's a lot of great companies that (00:01:12) are serving very important needs who are (00:01:13) well positioned to integrate um, AI into (00:01:16) their uh, products and services um, (00:01:18) because they either own the customer or (00:01:20) in the payment flow or what have you. (00:01:22) Um, and others that might struggle a (00:01:24) little bit more. Um and so I I think the (00:01:26) answer is very much it depends and both (00:01:28) to your question (00:01:31) >> and uh enterprise software has (00:01:33) traditionally relied on high switching (00:01:36) costs and recurring uh revenues. Do (00:01:39) these new AI capabilities threaten that (00:01:43) competitive mode more than previous (00:01:46) technological cycles did? (00:01:49) >> You know, I think that's a really good (00:01:50) question and I think it depends on what (00:01:53) is the moat. Um sometimes the moat was (00:01:57) uh oh we have this like network of um (00:02:01) data and we take data from here and (00:02:03) there and you're like in the workflow uh (00:02:04) tools. I think that is much easier for (00:02:07) an AI system to automate to understand (00:02:09) what is the workflow and and and change (00:02:11) it. That's fundamentally different, I (00:02:14) think, than other types of network (00:02:15) effects where um you're bringing a bunch (00:02:17) of people together in one place uh to (00:02:20) transact uh more efficiently like you (00:02:22) have with marketplaces or there's a (00:02:25) unique set of data that is built into (00:02:28) whatever ecosystem um that you own (00:02:30) that's actually much harder for um a (00:02:33) generally trained uh foundational model (00:02:35) to be able to replicate. Um and so I I (00:02:37) think that where um most will be eroded (00:02:40) I think will will really depend. Um you (00:02:42) know one of the classic you know I'll (00:02:43) just give you an example like every week (00:02:44) I think there's new new advances in AI (00:02:46) and so every week I think in some ways (00:02:48) my my opinion shifts but people talked a (00:02:50) lot about how CRM might get totally uh (00:02:53) shifted in the way uh by by LLMs and the (00:02:55) way you might have relational (00:02:56) intelligence. Um, and and I've believed (00:02:58) that for a long time. And also, uh, (00:03:01) arguably some of the best best (00:03:02) positioned companies to take advantage (00:03:04) of that are those that have the (00:03:05) relationship data already that are the (00:03:07) CRM. And so, who's going to win the war? (00:03:09) I I think it also depends. This will be (00:03:10) this will be figured out on the (00:03:11) battlefield in in the coming coming (00:03:13) months and years. (00:03:15) >> Um, Alexander, around $300 billion in (00:03:19) market value has been erased in a matter (00:03:22) of days as we have seen. uh how do you (00:03:25) separate real disruption risk from (00:03:28) indiscriminate selling across the (00:03:30) software se sector? (00:03:32) >> You know I am a venture capitalist and (00:03:34) not a not a stock trader and uh and and (00:03:37) the result is you know when when I (00:03:38) invest in a company my capital is locked (00:03:40) for five to seven sometimes 10 years and (00:03:43) so I I think much more in in in that (00:03:45) type of time frame than I do you know (00:03:47) two or three days. um the old adage of (00:03:51) um it's easy to uh overestimate what (00:03:53) you'll do in a year and underestimate (00:03:55) what you'll do in a decade. I think this (00:03:56) is very much what's happening like the (00:03:58) these changes are important but for a (00:04:01) lot of companies it's not tomorrow. Um (00:04:03) it's taking some time and I think that (00:04:04) for those that are positioning very (00:04:06) intelligently today um to reinvent like (00:04:10) fundamentally rethink how they deliver (00:04:12) their product um and how they um build (00:04:14) their organization. I think they're (00:04:15) going to be very well positioned to to (00:04:17) continue executing. Um, and I think (00:04:19) those that don't u, you know, buy or (00:04:21) beware on those and I I I think I think (00:04:23) that is that is the challenge. Um, (00:04:26) I I I I think there's some really (00:04:28) interesting opportunities both in the (00:04:30) venture and and and the public market (00:04:31) sides. Um, but but it's there's also a (00:04:34) lot of noise. (00:04:36) Alexander, private equity and private (00:04:38) credit have built large exposures to (00:04:41) software based on on assumptions of (00:04:44) stability and stickness. Um, does this (00:04:48) episode challenge does uh those (00:04:50) assumptions in a meaningful way? (00:04:53) >> 100% they do. Um, (00:04:56) this is the same type of shift that (00:04:58) happened a couple times before, right? (00:05:00) Like we went from analog era to the web (00:05:02) that fundamentally changed assumptions. (00:05:05) We went from the web to mobile um and (00:05:07) all of a sudden we had a new delivery (00:05:08) mechanism and way to reach customers and (00:05:10) went from mobile to cloud. Um I think of (00:05:12) AI as another one of these waves that (00:05:15) are a dislocations where the steady (00:05:16) state is no longer current and I think (00:05:19) three things are happening at once. One (00:05:21) um every single public company wants to (00:05:24) and needs to uh have an answer on what (00:05:27) they are doing with AI, how they're (00:05:29) adapting their business model, how (00:05:31) they're uh increasing the efficiency. (00:05:32) And so I think they're they're looking (00:05:34) for solutions and an answer. Um I think (00:05:36) too all of a sudden it is cheaper than (00:05:38) ever for an individual at a business or (00:05:41) an entrepreneur to experiment with (00:05:43) ideas, build their own SAS and that's (00:05:44) only getting easier if you look at some (00:05:46) of the things that cloud code released (00:05:47) recently for instance that powers some (00:05:49) of this in cloud co-work. I think we're (00:05:50) going to see a lot more of kind bespoke (00:05:53) especially workflow tools certainly you (00:05:55) know I'm I I run a small VC fund focused (00:05:58) on global innovation trends and um and (00:06:01) we are seeing that and we are (00:06:02) experimenting with that um candidly and (00:06:05) then I think three I think there's (00:06:06) genuine new value prop that is being (00:06:09) created that wasn't possible before in a (00:06:10) preAI era so yes like I think um I think (00:06:13) the old assumptions on that are um (00:06:15) getting rethought and um I should make (00:06:19) one last comment I actually think (00:06:20) there's a lot of uh experimentation. So (00:06:23) yes, everyone's trying AI uh and some of (00:06:25) the new tools and uh people are still (00:06:29) figuring out what is the right thing for (00:06:30) their business and whether or not and (00:06:31) where it adds value. (00:06:34) Alexander, as uh AI investment costs (00:06:37) rise and and cloud growth slows, um are (00:06:42) markets starting to shift the AI (00:06:44) narrative from growth potential toward (00:06:47) pressure on margins and and return on (00:06:49) capital. (00:06:51) >> You know, it's interesting. I um I wrote (00:06:53) a book a couple years ago called out (00:06:55) innovate. How global entrepreneurs from (00:06:56) Delhi Detroit are rewriting the rules of (00:06:58) Silicon Valley. It came out with Harvard (00:06:59) Business. It's in Spanish uh and a (00:07:01) number of other net languages. And in (00:07:02) the book, I talked about how the best (00:07:03) entrepreneurs outside Silicon Valley (00:07:05) didn't just focus on growth. They (00:07:07) focused on a uh portfolio of growth, but (00:07:11) efficient growth, strong unit economics, (00:07:13) profitability. I think that's really (00:07:15) what um is going to matter in this era (00:07:18) is being able to prove that you've got a (00:07:21) great long-term enduring business model (00:07:23) um that can scale that ultimately is (00:07:25) providing real value to the customer. (00:07:28) Um, and I think those businesses that (00:07:30) aren't just providing a, you were (00:07:32) talking about enterprise and things like (00:07:33) that a second ago. They aren't just (00:07:34) providing an enterprise service or a SAS (00:07:36) workflow tool, but that are closing the (00:07:38) loop on the service they're doing. You (00:07:40) know, you're offering a medical product (00:07:42) and you're you're having having help of (00:07:43) AI on the back end for a medical clinic (00:07:45) or you're offering a financial service (00:07:47) and and there's AI is part of that (00:07:49) story, but you're closing the loop by (00:07:51) actually delivering value. I I think (00:07:52) those types of businesses are still (00:07:54) going to have um a really good (00:07:57) positioning. Um and I think if they mix (00:07:59) that with like strong unit economics and (00:08:01) and and a great cash flow position, I (00:08:03) think that's that's what people are (00:08:04) going to be looking for and certainly (00:08:05) what I'm seeing at the early stage. (00:08:08) Um, Alexander, it's been a pleasure to (00:08:11) have you with us and to to go deeper in (00:08:13) in this topic which is very interesting (00:08:16) and our audience I'm sure are pleased to (00:08:19) to hear your your tips and everything. (00:08:22) So, thank you Alexander Lazero, managing (00:08:24) partner of Fluent Ventures. Thank you. (00:08:27) Have a good rest of your day and good (00:08:29) and good um weekend. (00:08:32) >> Pleasure joining you from Silicon (00:08:33) Valley. Thank you for the time. going to (00:08:35) blade.