Accel Sees AI Opportunity Shifting to Applications

The AI spending boom is showing no signs of slowing down. But as billions pour into models, chips, and data centers, investors are asking where the biggest returns will actually be. Joining us now is Matt Weigand, partner at Excel, focuses on late stage tech investments, including AI infrastructure. Mac and Debbie on the program, big big, I think, is fair to say to a lot of people. $3,500,000,000, this fund raised globally. Just a couple of months ago, back in August, actually, for emerging global AI startups. I think a lot of people would know your investments in Anthropic, Cursor, Perplexity, and the like. What are you seeing as the opportunities on the application layer in this cycle? Yeah, Tim. First, thanks for having me on. It's great to be back with you guys. Excel's been investing for the past forty years across the technology ecosystem. We've seen many of these cycles, and this AI cycle has us incredibly excited. We started out investing heavily in infrastructure then in models. And as you ask about applications, we think the last mile of delivery for applications presents an amazing opportunity for companies like Ligure on the legal side or even Cognition on the coding side. So translating the power of these models into actual value to the enterprise, we're seeing just a a number of really interesting emerging technologies. You know, I wonder what you look for as a venture capitalist in terms of of moat with this with this type of with the application layer. Because if if people are using maybe the same underlying models to provide similar services, how do you know which company will be the one that will be able to keep, I don't know, competition away or will be the one that actually will win in it? Yeah. Tim, when you think about knowledge work, there's the ability to augment in knowledge work and to provide value more in a a search based experience, many much of the way that consumers interact with AI today. But when you think about actually automating away tasks inside an enterprise, auditing auto automating away enterprise, knowledge work, that's where the last mile really matters. And incumbent in that is also permissioning and, rules based access controls for a lot of these technologies. This is the ditch digging that a lot of the application companies that we're investing in are are focused on today. You know, you mentioned that Excel's been around for four decades. You've been at the firm going back to 2013. We can, I guess, kinda think of that as, like, web two point o and and really the rise of of the application and and the mobile app and a lot of what we've, you know, seen technology built on over the last sixteen years in Silicon Valley? When you talk to folks who have been at the firm for for forty years and seen many different cycles, what's the context that they can provide about what this cycle looks like? Like, is this the biggest thing we've ever seen coming out of Silicon Valley? Absolutely. It's definitely the biggest thing we've seen coming out of Silicon Valley, And we've seen these cycles build upon each other and become more and more important with each incremental cycle, and they're always disorienting when they start. And there's a lot of questions. There's a lot of hypotheticals about where the technology will go. But if you just look at and I saw a stat just this week that across SpaceX, OpenAI, and Anthropic, you're looking at potential economic value that exceeds forty five years of technology IPOs leading up to this point. And that's just in the first four or five years post a chat GPT moment for AI. So we're still incredibly early, but we're seeing growth, like we've never seen before. Is there a crowding out that happens with the biggest players? SpaceX is now public, but with Anthropic and and with OpenAI because they have raised so much money and the valuations have been for private companies just been so extreme. Well, I actually think that's the fantastic thing about this technology cycle is you do have early leadership at the frontier for AI capabilities, but what you're seeing now is a huge ecosystem of open providing value to enterprises and applications, particularly around cost, sovereign data, and overall control of the model. So I think we're gonna see a a rising tide here that's gonna lift both the fronts front the frontier models as well as overall open source to allow enterprises to adopt this technology. You said control of the model, so I think it's a good opportunity to talk AI safety. It would be remiss if we didn't talk about this a little bit. I mean, this is sort of the main theme that's that's coming out of technology in in the UN right now at the UN General Assembly, and we saw Sam Altman at the White House last night at the state dinner with Xi Jinping and president Trump. I'm just wondering how you view a question that to many people has become existential about this technology. Is it really a threat to humanity? Look. I think what we're seeing from the frontier models is progress around that capability accelerating, which is fantastically exciting, but it also means that alignment needs to keep pace. The language you're hearing from the frontier right now is that we need to put more focus around alignment. And I think that's absolutely correct. Any technology that is being invented, whether it be the automobile, it be cloud, and shifting workloads to the cloud, required a commensurate investment in alignment to ensure that we're secure. And that's a responsibility not just at the frontier, but companies like CrowdStrike in our portfolio are thinking deeply about this. So I think we're gonna see an ecosystem of cybersecurity as well as AI companies meet the moment to ensure that we can take all the benefit of this technology and leave behind some of the risks that we're talking about today, the hypothetical risks that we're talking about today that could emerge. You you have got, I mean, Excel is like the poster child for a global venture capital firm, and we're talking offices in Silicon Valley, London, Bangalore. This new fund announced last month includes expansion for funds targeting The US, Europe, and Israel, and India investments. Do you think we need some sort of international framework of rules or regulations or standards for AI? I think the most important technology companies in the world of AI are gonna be the ones that develop trust with their customers, and and that's point one. At the end of the day, regulation will arise and always does arise for every technology that comes to the market, and I do think that there will be alignment frameworks and regulation that does show up. But I think first and foremost, if you want to build economic value and real value for your end customers, you need to develop that on the back of trust. And so the winners in this era, the companies that we care deeply about and spend time with are the ones that put trust first. They're not releasing something that they think is going to cause harm to society. They're not releasing something that could potentially be a cyber risk. They're building trust. Is that so that's a metric now when you're going out and looking at all the companies across your desk. That's a metric now that you're considering for investment. Absolutely. We back founders, and we care deeply about the culture and values of the companies that we back. And so we spend a lot of time thinking about what a future state looks like for their company and the ecosystems they play in. So it's of paramount importance to us when we partner with these founders.