Recent advancements in A.I. have everyone from programmers to journalists questioning whether they’ll soon be displaced by machines. But what about the very people funding the development of large language models—venture capitalists? James Currier, general partner at early-stage venture capital firm NFX and an A.I. investor, writes today’s Term Sheet and lays out which investors he believes are replaceable.
We are in the final 10 years of venture capital as we have come to know it. A.I. is going to remake the startup industrial complex, from its core. Venture firms will have to remake themselves into a combination of people and A.I.
Let’s be honest: Much of what a typical venture capitalist does—reading, summarizing, and ranking—is what large language models already do extremely well. Growth-stage VCs will be affected first. That’s where decisions are already made based on data, so it makes sense for A.I. to take over. Seed will be affected last: There’s less data available. Intangible human elements matter most.
A.I. will ultimately level the playing field among investors, making the market more efficient and lowering alpha, as software has done in relation to stocks and bonds over the last 40 years. But as we’ve seen with hedge funds, the best VC firms will get further ahead with a combination of A.I. and processes that give them sustainable information and operating advantages.
Three-person unicorns
A.I. will also change things for startup founders. With the next generation of A.I. tools, it’s plausible that three very talented people could get to $100 million in revenue with automated workflows. For software-centric companies, after an initial setup by humans, A.I. can be used to replace sales, customer service, a lot of the product design and coding, bookkeeping, and legal contracts. Vision and project management will be all that’s left to do. And with fewer employees, there’s a lot less time spent hiring, firing, and managing people. So that’s what startups can aim for: lean, three- to five-person teams run by aggressive, visionary generalists. A.I. can do the legwork.
These founders will need smaller checks from VCs, have fewer employees, and will move faster than ever. They’ll be able to do (a lot) more with (a lot) less. It’s also why the days of the “startup industrial complex” may soon be behind us, and why seed-stage VC investing will prevail in alpha going forward.
So what is the startup industrial complex?
It started in 1994 when browsers made room for unprecedentedly profitable business models such as online advertising and marketplaces. For instance, Facebook made $3 billion in EBITDA per month in 2021, and Google made $5 billion in EBITDA per month. These companies benefit from network effects, where products become more valuable as more users adopt them, same as SaaS businesses, with their own incredible margins and defensibility—when integrated into products and workflows, they’re hard to replace.
This made it possible for more people to be successful founders, and for more people to be successful VCs. Prior to 1994, founders in the tech space were rare: You had to be a bit crazy to bootstrap your own company. At that time, there were maybe 40 active VC firms and 150 partners investing in startups full-time. VC was a low-profile cottage industry serving outlier weirdos.
Today, the internet has made it easy. Now you can find customers, sell to them, charge them, and iterate your software product from the comfort of your office, without regard to store hours, geography, or the pesky handling of atoms. Money, power, and status have flowed to tech startups, and as a result, everyone wants to be a founder—and everyone wants to be a VC. There are now hundreds of thousands of tech startups per year and tens of thousands of people investing in tech VC nearly full-time. (We have 30,000 investors in the investor network we built at NFX, Signal.)
All these founders and VCs need conferences, news, education, pundits, legal help, community forums, gossip, and benchmarks. Welcome to the startup industrial complex.
‘Peak VC’
The startup industrial complex grew fat on itself. It might be entering its declining years in terms of producing alpha. Not only has the competition increased by orders of magnitude, but software and data are already making the market more efficient. There are more investors looking at more available digital data on startups from sources like Pitchbook, DealBook, Crunchbase, CB Insights. The patterns of what makes for a great tech startup investment are becoming widely known due to blogs, conferences, accelerators, and angel investing courses. And now there’s A.I. to accelerate all of that.
We could hit “peak VC” in the next few years, in terms of the number of people working in the industry, and then A.I.-aided competition will begin whittling that down.
The countervailing motion, which could offset it, is having every business move to software and tech, and the markets simply get bigger and more global. It will be interesting to watch the trends play out—and which ones win.
The age of A.I. will put an even greater emphasis on the outlier performances of the remaining humans—people capable of unique insight at ridiculous speed, people prone to strokes of genius, people who are, quite possibly, a bit odd. Their ability to build something great is only enhanced by A.I., and they’ll be able to manifest incredible value—even as the startup industrial complex winds down.
James Currier
Billions on the line…New documents obtained by Bloomberg News (that the FDIC didn’t mean to leave unredacted) reveal who Silicon Valley Bank’s 10 largest customers were and how much they had on the line when the bank collapsed. Sequoia Capital had a heaping $1 billion at the bank when it went under, and crypto stablecoin startup Circle had a balance of $3.3 billion. Mubadala-backed Altos Labs and fintech Bill.com both had hundreds of millions. Read the full story here. —Jessica Mathews
Jackson Fordyce curated the deals section of today’s newsletter.
