- 7 hours ago
Category
🗞
NewsTranscript
00:00By the end of last year, you invested in 165 different AI companies, so you are at ground zero.
00:06You know this better than anyone. There is huge anxiety that we are about to see tectonic shifts
00:11coming from AI. Just let's gut check from you. How disruptive of a force are we actually about
00:17to witness? We think AI is one of the biggest technology changes we will experience in our
00:26lifetime. People talk about the exponential rate of intelligence progression. And what that means,
00:34we have a saying, a little and a little and then all at once, is that people are experiencing that
00:39now as we hit the steeper part of the change curve where there's just rapid and very dramatic change
00:46because of artificial intelligence. And it's happening very broadly. It's 10x faster and 10x
00:52broader than we've seen with other platform shifts in technology. So I think that creates
00:57a question in people's minds of how do I adapt to this? How does society adapt when you have these
01:04changes happening so rapidly? And that is everything from questions about trust. If I have AI doing
01:10critical jobs for me, what's the technology that I can use to ensure that this highly intelligent
01:16computing platform is doing the things that I need it to do? It has to do with displacement and
01:22people have questions about the work that they do. If you are a company, who should be worried that AI
01:28is going to come and disrupt them? Well, this 2026 is the year AI is going to go from really
01:36demos in
01:37large organizations to big deployments. The market for pricing around that, it's this idea of agents.
01:42In the past, AI really was used only as a chatbot. Now agents act autonomously. They're handling
01:48everything from legal workflows, financial triage, software generation. Organizations are building
01:55bespoke products around AI. So I think what we're seeing is in any organization with knowledge workers,
02:02there's real questions. The speed of this is causing people to wonder how much of what they do today
02:08could be done by AI. And if that's the case, what do I do? What is my next role? And
02:13so we're
02:14companies are working through that. It's just the speed of this change. The exponential rate of change
02:19is, I think, caught everyone a little by surprise. It certainly has. And again,
02:22it's this moment of uncertainty and trying to dig it out. One of the companies you were a very early
02:27investor in is part of this change. Anthropic, I think you've been investing in it since it had
02:31less than 100 employees 2023. Just the latest with them, because this is a very competitive market,
02:37dropping their kind of hallmark safety measures. What was your take on that? Do you think that this
02:42is the right move for Anthropic? Well, first of all, it's an incredible company. It's in a 26-year
02:49lightspeed investing career. This is unprecedented. This is a company that grew from zero to 14 billion
02:56in revenues in less than three years. And this is really a testament to, again, the speed with which
03:02exponential speed of artificial intelligence progression and what's happening generally
03:08with AI. And now I think Anthropic in particular has been highly focused on this idea that trust
03:15technology, safety technologies, reliability technologies, as AI becomes more mainstream,
03:20those technologies have to exist along with artificial intelligence itself. Or we really
03:27won't have a way to let these systems go be highly energetic and doing the things that they would do
03:33independently in the world and for the world to have confidence in that. So I think they are deep
03:39technologically understanding those adjacent technologies. It's core to their mission. And I
03:46don't think there's anything that has changed about that. Now, the realities of how you deliver
03:52those capabilities into the market, I think, you know, it's such a dynamic market that there are,
03:57you know, evolutions in how they think about, you know, exactly how they deliver. But the core mission
04:02and the core technology advantages we believe they have in trust and safety so that we can have really
04:07confidence, you know, so that an AI system, you won't want it as a demo. You really want to ship
04:11it in
04:11production because you can know for critical application, it will do what you need to need
04:16it to do. We think they're far ahead in that domain. Ravi, I mean, yesterday was a great example
04:20of you. You keep hearing of these different partnerships that Anthropic is dealing, is
04:24stealing and all the companies, it seems like big beneficiaries. But what exactly is the business
04:30model? Where does the revenue come from? Is it partnerships? They get money from that. Do they
04:35get the data from it? How does that exactly work out? Right now, for all intelligence model companies,
04:44the revenue model is by letting the intelligence tokens be used by different applications,
04:52by end users. And there's there's a strong monetization model that I think in Anthropic's
04:56case, because they are so focused on the enterprise and our large institutions, public institutions in this
05:02country. The idea that you could create digital co-workers around their, you know, highly intelligent
05:07models means that you have also an opportunity to to price and have monetization around the idea that
05:15what used to be SaaS software now can be done by a digital agent. The core workflow products, you could
05:21just
05:21have a general digital co-worker that performs those duties. And so there's a lot of different ways that value
05:28will be will be ascribed to Anthropics technologies. It could be product. It could be SaaS replacement.
05:33It could be the fact that the work in an enterprise or a large government organization that has real value.
05:39And and now the models can offload some of the some of the lower level, you know, kind of repeat
05:44knowledge work.
05:46So, Ravi, I mean, you have some pretty big investments in foundational companies. Again, Anthropic,
05:52XAI, it's not every VC that has the scale, the size, the connections to be able to invest in companies
05:59like that. What happens to the rest of your peers? Is there going to be a big shakeout if you're
06:05not here
06:06and forward with the companies that are leading these changes?
06:12You're right. And Lightspeed as a platform has evolved to, we think, meet the needs of the
06:18the the A.I. technology generation. We raised nine billion in our across six funds recently.
06:24And the you know, we're more than one hundred and sixty five A.I. native portfolio companies,
06:29nine of the top 20 most valuable companies globally in A.I., which is more than any other firm.
06:34I think our view is that the capital industry to support artificial intelligence
06:40tech companies, it has fundamentally changed. It's raising capital or investing in these companies is
06:46five to ten percent of what we do now. There's all kinds of other things as an investment, a VC
06:52investment
06:52platform you have to do when companies grow from zero to tens of billions in in just a few years.
06:57It's about helping navigate regulatory frameworks, strategic partnerships, all the chessboard planning
07:04across global markets. So those things are are critical, not just providing the money. And you have to have capital
07:10at scale with real staying power. That's going to be, you know, both the public markets where, you know,
07:16investors are chasing quarterly returns. And you're right. There's a need for scale to be able to stay with
07:21these companies and help them all the way through to the other side. It just looks different than, you know,
07:26what venture capital used to be. So how many have been able to keep up with that? In other words,
07:29Ravi,
07:30how many kind of zombie VC funds are out there wandering around, unable to adapt and just
07:36babysitting or winding down their current portfolio? Well, I think the other thing, you know,
07:41when you take a company like Anthropoc, when that goes from first dollar to over 14 billion
07:45revenues in under three years, it does pay to be early. And, you know, there's an opportunity,
07:50we think, for early stage funds if they can understand, you know, again, where the trend is
07:55going, where the super cycle is going, not the current generation of technology because of how fast AI is
08:01moving. There's opportunities to invest in things that can grow and become extraordinarily valuable,
08:06but you have to be there early. What are those things, Ravi? Because it seems like at the moment
08:11that the foundational AIs are the ones that are eating the world right now. Again, we think the 2026,
08:19the first year we're going to see, you know, AI in large deployments. This is in large enterprises and
08:24public institutions go from demos to deployments. And that's really about agents and the fact that agents
08:30can really go from being, instead of chatbots, these things that do long independent complex
08:35work. Actually, if I can jump in just on that point, does that also mean that 2026 will be the
08:40year that we start to also see mass layoffs tied to that? If you have AI agents that can all
08:45of a
08:46sudden do this work? So far, what we're seeing is the agents let people be more productive because
08:53they can offload a lot of the repetitive tasks. And so there's a, there's enough leveling and most
08:58companies. And this is, again, I think traditional enterprises is happening so fast. A lot are
09:03thinking how we use AI to be 10% or 15% more efficient, but the rare few that are
09:08really
09:08thinking if I had unlimited access to intelligence, how would I reimagine how my entire company works
09:13in those companies? We're not seeing that, uh, jobs go away, but we are seeing they have an
09:20opportunity to take the people they have and do much more, whether it's add features faster,
09:24whether it's be, you know, more, uh, uh, more broad in their marketing coverage because
09:29people use these tools. So we're seeing that the, the, the most AI red pill, so to speak,
09:34companies are able to grow their market position significantly. And they're looking at that with
09:41the people they have. Now there is some, I think reallocation, a lot of the lower level work that is
09:46repetitive in certain parts of the knowledge work broadly in that domain for businesses is, is going to,
09:53you know, go to the AIs. And I think the key is going to be people have to adapt. They're
09:57going
09:57to have to learn how to exist in a world where you have both artificial labor and human labor,
10:01you know, kind of existing side by side. So I think what you are seeing is this kind of real
10:06push
10:07for people to have to learn a new way of working, given how much these agents, these AI agents are
10:12capable of doing. And that, you know, could cause some shifts where if people aren't able to learn and
10:17and skill retrain, you know, then the old opportunities that they had.
Comments