My latest at Forbes:
Will the home be the last place the humanoid robot gets a job? According to Sanctuary AI CEO James Wells, probably yes — and that’s a bit of a gut punch for those of us hoping robots will soon be doing laundry, dishes, and cooking.
I caught up with Wells at Web Summit Vancouver. Sanctuary is Canada’s only homegrown humanoid robotics company and holds what Wells says is the third-largest IP portfolio in the space globally. What he told me just might reframe much of the current humanoid hype cycle.
The home humanoid timeline: “three, five, maybe seven years.” By every axis Sanctuary uses to rank deployment environments — unit economics, environment complexity, customer sophistication, safety tolerance — the home ranks last. When I asked Wells about 1X’s Neo, which is being pre-sold for home deployment now, he chose his words carefully: “I applaud their marketing initiative. Which is a marketing initiative.”
The core problem: “In the industrial world, you need to be 99.999% repeatable. Most of these foundation models get you to about 80% performance. So you can do a lot of different things, but not that well. So you’re dropping a glass one out of five times.”
Wells argues that hands — not legs — are the real gating factor. “Hands are the gating factor for physical AI to proliferate into the world.” Sanctuary has specialized in hands since 2018, taking a super-contrarian hydraulic approach while the rest of the industry went tendon-driven and electric. Their hydraulic valves are coin-sized, food-safe-oil-actuated, and tested past two billion cycles without degradation — 50x faster and 6x cheaper than off-the-shelf components, the company says.
Then there’s the sovereign labor question. Humanoid robots aren’t just products — they’re labor, and that means they’re GDP. Wells recently spoke with Canada’s first-ever minister of AI about it: “If you do nothing, you will be forced to buy Chinese robots with AI brains that Canadian business will hire and you will hollow out the entire economy.”
As for the iPhone moment for humanoid robots — the ChatGPT moment when a massive phase shift becomes undeniable — Wells thinks there won’t be just one: “There’s going to be moments along the way. Task by task. Unlock a group of tasks, unlock another group of tasks.” The endgame is zero-shot learning: a robot walks into a brand-new situation and immediately starts doing useful work. We’re not there yet. But we’re moving forward quickly.