Happy Sunday and welcome to the Investing in AI newsletter. I’m Rob May, a Partner at PJC. Don’t forget to check out my podcast interview with Bloomberg Beta Partner James Cham, and most recent one on synthetic data with the Synthesis.ai team.
This substack covers investing in AI, and topics related to that. When we get technical, it’s to explain how that technology filters up through products, business models, and competition in markets.
— Interesting Links —
Baidu’s Chip Unit Raises A Round. Reuters.
Gartner’s Magic Quadrant Cites a Glut of ML Innovation. Venturebeat.
The Pastry AI That Learned To Fight Cancer. New Yorker.
It’s Time For Global Treaties on AI. Brookings.
AI: Are We Doing It All Wrong? CNET.
— Commentary—
U.S. Steel was the first billion dollar corporation, in 1901, shortly after it IPO’d. Obviously a billion dollars nominally was worth a lot more compared to one billion today. Then it was 54 years before the first $10 billion dollar market cap company - GM in 1955. From there, IBM crossed $100 billion in 1987, then Apple crossed the $1 trillion mark in 2018.
As you can see, the numbers go up by an order of magnitude, and do so at a faster pace, although a pace that is somewhat gated by the size and growth of the world economy. That makes me wonder… when will we see the first trillion dollar IPO? It’s probably closer than we think, driven by several factors, including inflation, more startups growing faster, tools that make it easier to start and scale, and automation coming from AI. It’s possible AI will enable the first trillion dollar IPO.
The question is - what will it look like, and how could you identify that company today if said IPO is 12-20 years away? Here are the properties I think you look for in that company.
It will be in a brand new massive market. This is actually the point I’m least comfortable predicting, but I’ll say it anyway. Google and Facebook were new markets. Uber, another large tech IPOs, was sort of replacing and eating into taxis and other forms of transportation. It really depends on how strictly you define “new market.” But when looking at companies that could grow into a $1T market cap, you have to imagine a world where some new device, service, experience, etc becomes so common that almost everyone in the world does/has it.
It won’t be an obvious extension of existing tech. If it was, the big tech companies would do it. This will be something new that will come up on them too fast, or, run counter to their business models and thus make it tough for them to do.
It will make incredible use of AI powered automation. This will help it scale fast, with fewer employees than other companies did through similar growth curves. In my opinion, this is the area where companies are the most behind - applying AI to common workflows.
It will be built on more of an applied AI breakthrough than a theoretical one. It’s really rare for pure new tech companies to get big fast.
It will be profitable quickly. This is one thing that is changing about AI tech. While seed rounds will stay the same size, I think growth rounds will be smaller as companies need less capital to scale.
For a long time now I’ve believed that the window of 2018 - 2028 is the decade in which a company (or two) like this will be started. While the future is hard to predict, these are my thoughts on what a company should look like, conceptually.
Thanks for reading.
@robmay