everyone is building for a room you will never stand in

almost every humanoid robot with real money behind it is being built for a room you will never stand in. a warehouse. a fulfilment aisle. a factory floor at three in the morning. the pitch writes itself, the buyer has a budget, and nobody has to like the machine.

i talked to a founder this week who has been at this since before the category had a name, and he went the other way. he is building for the living room.

his reasoning is not sentimental, which is what surprised me. a machine in a factory has to hit a cycle time and not hurt anybody. a machine in your kitchen has to do all of that and then be something you want in your house. it has to move without that lurching stutter that makes your shoulders rise. it has to read the room. the technical bar is higher, not lower, and the commercial bar is stranger, because the buyer has to have feelings about it.

he picked the harder one on purpose. he wants to build something people have emotions about rather than something nobody ever sees.

the better story from that conversation is older. years ago he built a showpiece. a full sized machine that talked to strangers and did a real physical thing on command, wired to real hardware, and it stopped people cold at a conference. crowds. attention. the works.

it was never going to be a product. the setup was expensive and fragile, it needed gear nobody would maintain, and in a loud room it could not hear a word anybody said. the demo that wins the room and the machine that survives contact with a real house are two different machines, and he learned that with his own money.

then the part i respect most. his original architecture, the thing he founded the company on, he threw out after two or three years because the hardware and the algorithms were not there yet. mission kept, thesis binned. most founders do the reverse and defend the thesis until the money runs out.

that is the actual lesson under all of it, and it is not about robots. the bet everyone can see funded is the safe one, which is why it is crowded. the interesting bet is the one where the technical problem is harder and the buyer has to feel something. and whatever you picked, the demo that draws the crowd is almost never the thing that ships.

the machine that works in a loud room is worth more than the one that wins the stage.

so which of your bets is a demo, and which one survives a loud room?