On Y Combinator startups and the AI revolution
ChatGPT launched on November 30, 2022, and gained over one million users within five days. Now a couple of months later, it’s fully mainstream with over 100 million users — this is faster mass adoption of a technology product than I’ve ever seen.
It’s easy to forget that OpenAI launched ChapGPT as a ‘research preview’ fresh out of their lab for early beta testing. It wasn’t meant for mass adoption, and Sam Altman (OpenAI’s CEO) said, “it’s a mistake to be relying on it for anything important right now,” but that didn’t stop millions of people from using it every day. This is the very definition of product/market fit.
At Rebel Fund, it’s our job to invest in the top 5–10% of seed-stage Y Combinator startups each year, and since AI has recently taken the technology world by storm, the goal of this post is to share our thoughts on where AI is and where it’s heading.
As usual, I’ll start with a chart:
Pulled from our database, this shows the percentage of new YC startups building AI products by year, including the startups that have publicly launched so far in the current Winter 2023 (W23) batch.
You’ll see that YC startups’ AI interest started around 2013, reached a peak in 2017 (see Gartner Hype Cycle for Emerging Technologies, 2017), fell into a mild trough of disillusionment in the succeeding years, and has reached a new peak this year, no doubt thanks in part to ChatGPT.
The big question I’m getting from the media and our fund investors is, “Is this just another hype cycle?” and my answer is “No”
AI has always had a certain inevitability to it, and scientists like Alan Turing have been writing about it for at least 80 years. The question has never been “if” humans would develop machines smarter than ourselves, but rather “when” and “what will be the implications?”
There were a series of recent technical breakthroughs that led to where we are today with AI¹, and there will need to be a few more to reach the holy grail of artificial general intelligence (AGI), which is the ability of an AI to understand or learn any intellectual task that a human can.
Even though we’re still far (maybe not too far) from developing an AGI, I think we’ve reached a tipping point in the evolution of AI that will make the next 10 years look very different than the past — and Y Combinator startups will be in a unique position to build and capitalize on the products that help bring AI into every aspect of business and life.
It’s no secret that Sam Altman was formerly president of Y Combinator, so OpenAI and YC have a special relationship. But whether OpenAI, Google, Anthropic, or others build the dominant AI platforms, it’s startups who will build the products and user interfaces that bring the power of these platforms to the masses².
Since I’m a venture investor and not a philosopher, I won’t speculate on all the ways that AI will change the world, but in terms of magnitude, I think it will be closer to “electricity” and “industrialization” than “mobile” or “blockchain” — and the AI revolution will happen much faster than any of these.
Plenty of individual AI startups will get overhyped, plenty will fail, and some will become billion or even trillion-dollar companies.
As with any other technology vertical, the most important things for investors to look at with AI startups are team, product, market, traction, timing, and defensibility — and those who get carried away with heat or FOMO will quickly lose their shirts.
That said, the power of AI should not be underestimated, and I’m excited to meet and invest in YC startups building the most promising AI-powered products in the years ahead.
¹Despite all the hype, today’s LLM-based models are really only good at text manipulation and generation, which on the one hand, is a pretty narrow use case, but on the other hand, represents a good chunk of white-collar work
²It’s a curious twist in history that OpenAI developed ChatGPT, as a startup could have built a similar chatbot on top of GPT-3 well before. Oops!