Depending on whom you ask, artificial intelligence will either vastly improve our lives or take away our jobs. For veteran venture investor Peter Wagner, it’s a little more nuanced than that.
We recently caught up with Wagner, who, along with fellow veteran investor Gaurav Garg, launched Wing Venture Capital. Combined, they have upward of 25 years of experience at storied investment firms: Wagner joined Accel as an associate in 1996 and stayed more than 14 years before leaving as a managing director to co-found Wing, and Garg spent 11 years as a partner at Sequoia Capital.
“That was really unhealthy, the idea of coming back to raise a new fund every year or two. That type of time compression is not good for anybody.” Peter Wagner
Since launching Wing in 2011, the two have been steadily growing their team and portfolio, backing such companies as Snowflake, Gong and Cohesity, among others, and “helping people do their best work,” as Wagner told us.
We initially reached out to discuss a list of enterprise angel investors that Wagner and his team recently assembled for the sake of the enterprise investing “community,” but we wound up talking for a bit about the firm’s growing focus on AI startups in particular and some of Wagner’s separate thoughts on the current state of the venture industry.
(Editor’s note: This interview has been edited lightly for length and clarity.)
You’re investing in the “AI-first technology stack.” What does that mean?
We’re big believers that AI is not replacing humans, but that humans working with AI will replace humans who don’t work with AI. That set of types of products really empowers customers to do more and to reach their full potential. … The whole notion of helping founders and their customers do their best work through the application of AI-first technologies, that’s really my mission statement. And that is a lot of what is driving enterprise technology these days, the question of: How do I develop and apply that set of AI-first technologies to drive business results?
You think the next generation of business applications will be powered by AI. But are there risks? The Information wrote in late December about Jasper, a copywriting startup that’s built on OpenAI’s flagship large-language model, GPT-3. It noted that because of the strength of OpenAI’s free-for-now chatbot, ChatGPT, Jasper is being forced to find new footing. It seems we could also see the network effects of companies like OpenAI getting bigger and kind of crushing everyone else along the way, no?
Well, crushing them or enabling them, right? The foundation models, to use the term that OpenAI and others are putting out there, are super powerful. But what they do is create opportunities for startups to build around them and on top of them and add value to them. This makes those models more useful in specific use cases and more useful to specific customers. So I think of the amazing developments amongst these foundation models like the creation of the microprocessor or the creation of the PC. These were enabling technologies.
Did the emergence of the PC destroy industries or allow new ones to come into being? A little bit of both, but much more the latter. Think of all the companies and different types of software products and other industries that were born because of this enabling technology.
An outfit like Jasper, just to stick with your example — it needs to figure out how to add value around the OpenAI family of models, and if it can’t do that, you don’t have a company. I think what we’re seeing is many founders finding ways to extend and add adaptation layers and workflow around the foundation models to really drive business outcomes. And users will pay for that.
Presumably, you need fewer people because of the power of these models. Is that being reflected in the amounts of money that people are raising?
You certainly don’t need people going and recreating the work that OpenAI has done. If I go back to my chip analogy, it used to be if you were a computer company, you had to develop a microprocessor that was part of IBM and Digital and others that preceded Intel were doing. But when commercial microprocessors came out and became sufficiently powerful, the computer industry no longer was spending time developing microprocessors; they were developing a lot of other things that added up to a more complete and more usable product. So I think you’ve seen resources shift. [Startups will] take advantage of the enabling technology and then allocate their resources elsewhere. I don’t necessarily think it means that there are fewer people in the company; I think they’re doing different things. If that’s not the case, then it might be that their value add is relatively thin.
If not the amounts of money being raised, what is changing as we head into 2023? Are you seeing much at the seed and early-stage levels? Are you concerned about the economy right now? Do you think it’ll be a slow year or that things will be fine in another six months?
Well, I don’t think it’s going to be fine. I think we’re in for an extended bottoming period. I don’t know for how long, but if prior experience is any guide, we could bump along in a bottoming period here for a year or two. It doesn’t have to be doom and gloom, though. If you look back, many of the best companies of enduring value were formed during these bottoming periods.
I also think we will see deal sizes perhaps normalize a bit. The huge fundings were supply driven; it wasn’t that the companies needed that capital and a lot of that capital is still sitting on balance sheets.
Will the size of venture funds normalize a bit? Big funds got really big. Do you think these same VCs will raise less next time around, or else space out their fundraisings more than they have?
I think certainly there’ll be spacing out. You know, that was really unhealthy, the idea of coming back to raise a new fund every year or two. That type of time compression is not good for anybody. As for size, if you go back to the early 2000s, after the dot-com bust, we actually did see large fund size reductions. Not only that, but we saw capital that had been raised actually released.
You were at Accel when Accel did this, among other firms.
Absolutely, yeah. So we ended up releasing a large fraction of [Accel’s eighth fund], and other firms did the same. But what has happened now that’s different is you have managers on a multistrategy approach, so whereas Accel at the time was doing early-stage investing, Accel today is doing growth stage and a number of other different strategies, and I think you’ll have capital finding new strategies. I’m sure you’ve probably reported on how some managers were deploying into the public markets capital that was raised for their venture funds. I think that’s a bad idea. If I was an [investor in someone’s venture fund] I wouldn’t be happy about that. But I think the capital is more likely to be deployed in different strategies right now than it is to be released.