Schrodinger’s Opportunity
Our humanoid robot future remains in a state of superposition
By Patrick Hunt
Quantum physics has this crazy concept called superposition - when two or more mutually exclusive outcomes exist at the same time. These states continue to exist until someone observes the situation and they collapse to a single reality. There’s a famous thought experiment - Schrodinger’s Cat - in which a cat in a box is both alive and dead until someone opens the box to observe.
Superposition might seem pretty irrelevant to our daily lives, but we do see its analog occasionally in capital markets.
It’s typical for people to hold divergent views about an investment opportunity. As more information becomes available, they collapse to a consensus view. But sometimes views remain in superposition, even as information piles up. The divergent, mutually exclusive realities expected by investors harden and become more real for each observer. I call this a Schrodinger’s Opportunity.
Humanoid Robots in Superposition
Humanoid robots present a fantastic Schrodinger’s Opportunity. One set of observers is confident that the future of robots is humanoid. Morgan Stanley agrees, estimating that by 2050 we’ll have one billion humanoid robots in a market worth $5T! MS estimates a 'humanoidability factor' for each labor classification and models an adoption curve up to 2050. The typical argument behind this viewpoint goes like this:
Advances in software typically influence hardware form factors
AI is accelerating, and any bet on its limits will prove foolish
We’ve shaped the world around us; as software capability approaches human capability, robot form factors will approach human form factors
Figure, Tesla, and other humanoid robot companies will generate immense value
Another set of observers is confident of the exact opposite. They’re bullish on robots, but don’t think they’ll look like us. We’ve been automating work for centuries and the non-humanoid solutions typically work best. They argue:
Specialized hardware offers lower cost and higher performance; software won’t change that
AI is impressive, but we’re a long way from spatial AGI
In the built environment, wheels are better than legs, and heads are ornamental
The value in robotics will accrue to verticalized players who specialize
When we get more information, like videos of Figure robots in BMW’s Spartanburg plant, these views don’t collapse but instead harden into more confident opposing realities. You can practically hear the debates unfold:
Humanoid Lovers: See?! It’s already working. A humanoid robot can already do a person’s job. And this is just the beginning. The number of humanoid robots in manufacturing is going to keep growing.
Humanoid Haters: Look at the customized fixtures, racks, and welding robots. Everything in this factory is customized to the jobs involved – including the existing robots. This is a marketing exercise with no real value.
Doesn’t it remind you of a certain car company? Tesla is simultaneously one of the most valuable companies in the world and one of the most shorted stocks. Every earnings call provides new reasons to believe and more reasons to short. Tesla has been public for 15 years and we’re still doing this.
How to Think about a Schrodinger’s Opportunity
It’s tempting to adopt one side’s reality or another’s, but if we’re going to be independent thinkers we need to segment each side’s argument and form our own thoughts about each issue. As more information becomes available, we should reevaluate whether it supports or contradicts one of the argument segments. Here’s how I’m making sense of the humanoid robotics opportunity.
The Role of Hardware
The State of AI
The Winning Form Factor
Where Value Will Accrue
All the Boxes
The cat is alive. The cat is dead. The opportunity is in understanding that the cat isn’t the product; it’s her kittens, each a little different, every one of them in a different little box, hoping it will be opened by the right new owner and taken to their forever home. Many of them won’t.
Claiming “X is the future” yields zero insight into which subset of X is worth investing in. Claiming “X will never happen” is medium-term thinking that rejects the possibility any subset of X will ever find product-market fit.
Schrodinger’s Opportunity lies in rejecting binary thinking and the simplistic narratives that flow from it. Progress is inevitable, but success is not. Someone will build a successful humanoid robotics company, and we’ll all miss it if we’re too busy being right.
References and Credits
Thanks to Alex Roy and Joel Johnson for editing.
Thanks to Reilly Brennan and Will Foss for good convos about where robotics is headed.
If you don't know what Narrative Command is, read this.
If you're excited about robots inspecting water mains, check out Motmot.