We got Keith in the waiting room. Oh, yeah. That's perfect. You have to imagine Blue Ghost was running on ramp to make it to the moon.
Oh, undoubtedly. You can't get to the moon if you're taking 45 days to close your books every month. It's impossible.
What do you think, Keith? Is it possible? The future is on ramp. So all the companies that want to create the future hopefully are saving time and money on ramp.
Yes, yes, I love it. Can you give us the update? What's going on with ramp? They seem to be on a tear, valuation never been higher.
Actually, I would push back and say take a victory lap. I know the job's not done; Eric has been very clear about that. But I feel like you can take a little bit of a victory lap, and probably more this year to come given the momentum.
Well, the amazing thing about ramp is the value proposition has been clear from the beginning. We're going to save you time, we're going to save you money. And now there's empirical evidence across any type of company; we can show five percent this, ten percent that. So it's very obvious. The CFO world is one of the most archaic parts of any business. Even the most cutting-edge organizations typically have a very archaic CFO function. Ramp has the vision to digitize and bring the CFO function and all the finance into the modern age, using the same tools, techniques, data, software, and SaaS tools that everybody else in your organization uses. That's why it's such an exciting vision. The valuation is somewhat art, not science, but the reason investors across the globe are excited about ramp at any point is because we have this vision of transforming the CFO suite. Every company has this broken engine from the 1800s or 1900s, and we're going to bring the engine of finance into the 21st century. AI will probably bring us into the 22nd century with ramp too. So that's why everybody's excited.
The other thing that's special about the company in the short term is that usually at this level of scale, companies slow down. The product velocity slows, the quality of talent decreases, we have a reversion to the mean. But ramp is firing on all cylinders. The product velocity is increasing, the quality and insights of the product and the team are improving. The intern classes for the last two years are by far the best intern classes anywhere in the globe. So that's the future. That's why this round is very exciting; it's a prediction for the future based on a lot of savvy people who get access to all the best companies. Ramp still has the highest slope.
Do you think that's because Eric and Kareem are second-time founders? Can you break them down as founders, and what makes them able to stay on this insane cadence going into year six, going into year ten? How do they have so much energy?
Great question. I wish I knew the answer, because I would love to invest in more. If you can come up with a formula or predict a formula, that's great. But having worked with them now for over five years, I think there's a couple of ingredients. One subtle ingredient is Eric has amazing marketing instincts. You really can't train this. The best CEOs are the chief marketing officer for the company. Eric entered a space that was crowded, with large and funded startups, and he was able to cut through the clutter immediately. He deserves the credit for clarifying succinctly, powerfully, and communicating the value proposition of ramp, and then reinventing that every year or two to make it broader, more relevant, more impressive. Kareem is setting the engineering tempo and quality bar from the beginning. The combination is really interesting. If you look at some of the heroic companies in people's portfolios, like Stripe, you have two brothers, so you're getting two executives for one. With ramp, there are two CEOs for one. That incredible synergy between two incredibly talented people who are always on the same page and reinforce each other is very rare. So if I can find two other founders that are that much in sync, that might be the reason to invest.
What are you seeing for young Gen Z founders? There's this narrative that Gen Z hasn't produced their Zuck yet. What's your take on that? And what would be some advice for younger entrepreneurs coming in, maybe deciding between building a ChatGPT wrapper to make some money versus taking a big shot at something massive? How do you counsel people these days?
I don't think there's a formula that works for building an iconic company. As you point out, Zuckerberg's first company was Facebook. Eric and Kareem built a company and sold a company. So I don't think there's a one-size-fits-all formula. It comes down to whether you have a vision of what the world needs and whether you have a strategic advantage in your own traits and personal superpowers that maps to that problem set. If you have that when you're 19 or 20 or 21, it might be a really good idea to proceed. If you don't have that combination, then learning, building, working with other talented people might be a good formula. For sports fans, there's an epic debate in football about drafting a rookie quarterback and whether to start them or sit them on the bench. I've studied this empirically, and there are Hall of Fame quarterbacks from both backgrounds. Those who start right away typically have a miserable first season; those who sit for two or three years, when they start, they don't have a down interception-to-touchdown ratio. So it depends on the quality of the team. I think it's most useful for founders to work in an outstanding organization for a year or two before starting their own company, to get a feel for what great looks like. There isn't a great substitute for that. But if you're ready to go, there are people who learn better by being thrown in the pool to learn to swim.
I love it. My question is, you noted earlier that ramp was in a competitive category. Do you see categories today like fintech in 2020 when you made your first investment in ramp that are ripe for a team to come in and take a new approach, potentially invert the value prop the way ramp did? Have you identified anything, or are there categories you'd like to invest in?
I was fortunate in 2019. I had a fair amount of expertise in the space and concluded there's massive opportunity in replacing the Amex business card and the bigger vision of digitalizing the CFO's world. The people funding other operations were missing key ingredients and naive about underwriting. Unusually for me, I was proactively trying to find high-quality founders interested in this problem set. Delian eventually intercepted... He'd been listening to me pontificating every day about needing to find people to compete with Brex and the like. He walked into my office and said he found the founders. I was shocked, but he did. I don't usually do this, but I had a lot of expertise in the area, so it was non-standard for Founders Fund. The most acute point was the monopoly story: if all these competitors are doing well, what's the monopoly story? There is one, but pitching Founders Fund in 2019 in a category that included MX, SVB, Brex, and others, took a little massaging. Fortunately, Brian Singerman saw the potential and was alienated by the other founders too, which helped correct the antimonopoly crusade.
Are there any categories you're looking for now, hunting for founders? Obviously, everybody talks about AI. There are opportunities in AI that probably fit that characteristic where people have assumed a winner, but perhaps the right founders are now five to ten years into that story. Sometimes it takes time to bake before you get the next generation of founders with insights into what could be reinvented. Consider what SpaceX did: rockets were good but invented in the 1960s, and it took about 20 to 30 years before SpaceX reinvented them. Rocket Lab came another 20 years after SpaceX. So sometimes you let the wave go on for a while before you have a vision to reinvent it. I would love to invest in something that's a next-generation replacement for LLMs—not a better LLM, but something that transforms it and makes LLMs obsolete. That would be a cool investment.
How durable do you think AI revenue is today? You were around for the dot-com era, where companies went public with little to no revenue. The counterpoint today is that we have great private companies with nine-figure ARR, growing fast, but we've talked about easy come, easy go. If you can create traction quickly, how durable is that? Do you think companies adding $10 million of AR every month are durable and will be leaders in five years?
There's a quality of revenue. For example, OpenAI's revenue mostly comes from ChatGPT subscriptions; that's pretty durable. Consumers use it for business applications. But you can get revenues that are more like pilots, which are non-durable. In between, if you sell to an enterprise and they roll out the AI-based product to the entire organization, the chance they'll swap in the short term is very low. If you get through procurement and have a champion and change business processes in a large organization, that's durable for probably five plus years. Those AI companies are in good shape. But some adopt on a trial basis; that revenue is precarious. Another set of revenues revolves around training and post-training. Advances in AI are so fast that how we build models today with fine-tuning may be obsolete in a year or two. Revenues built on that kind of training may not be as durable.
While we have you, what's your take on humanoid robots? Everybody has a take, and I'm guessing you have something spicy.
The spicy take is that the Chinese are ahead of us, and it's very dangerous politically and geopolitically. The administration will have to take action. Large companies like Meta are publicly embracing it, and their CTO is responsible for this. Many startups are well-funded; it's probably the hottest sector among VCs. You walk in, and they ask, 'How much money do you want?' and it's hundreds of millions. Tesla has made progress; I think some private progress is better than the public stuff. So the sector will be successful for some. As a VC, it's hard to invest given the proliferation of startups with high potential and high prices. It's unclear if there will be monopolies, yet some are priced that way. We need to learn from DJI. They flooded the market with products sold below cost. Unitree is doing the same with humanoids. The Trump administration will have to address this. Robotics has military applications too. The data collection potential and military applications of a robot that can run 17 miles an hour mean this administration will figure out a way to ban Chinese robots before it's too late.
