Figure AI just raised $2.6 billion and the

Figure AI just closed a funding round that should make everyone pay attention.

$2.6 billion valuation. Investors include Amazon, Microsoft, NVIDIA, Jeff Bezos personally, and a list of other names that reads like a who’s who of technology. These aren’t angel investors making small bets. These are the companies that run the infrastructure of the modern economy saying “humanoid robots are real and we want a position.”

Bloomberg broke the story. The number is staggering for a robotics startup that hasn’t deployed commercially yet. For context, Boston Dynamics was acquired by Hyundai for $1.1 billion in 2020, and Boston Dynamics had been around for decades with a proven product line.

Figure is two years old.

Follow the money

I’ve learned to watch where the money goes rather than what the press releases say. Press releases are marketing. Investment checks are conviction.

Amazon operates the largest warehouse network on Earth. They already have 750,000 robots in those warehouses, little orange platforms that carry shelves. Amazon is investing in Figure because they see humanoid robots handling the tasks that shelf-carrying robots can’t: picking individual items from bins, packing boxes, handling irregularly shaped products.

NVIDIA makes the GPUs that train AI models. They’re investing in Figure because humanoid robots need AI brains, and those brains run on NVIDIA hardware. Every deployed humanoid robot is an NVIDIA customer.

Microsoft provides cloud infrastructure and AI models. They’re investing because the software layer for humanoid robots (perception, planning, language understanding) will run on cloud compute.

These aren’t speculative bets. These are companies investing in their own future supply chain.

The timeline I believe

I try to be conservative with predictions. I’ve been burned by robotics timelines before (Boston Dynamics has been “five years from commercial deployment” for twenty years). But the convergence of factors in 2024 is different from previous years.

Language models now give robots the ability to understand verbal instructions and reason about tasks. Computer vision has reached the point where a robot can identify and locate arbitrary objects in unstructured environments. The cost of actuators, sensors, and batteries has dropped consistently for a decade.

The limiting factor used to be intelligence. Robots could move but couldn’t think. LLMs changed that. Now robots can think but are still learning to move with precision. The hardware-software gap is closing.

My guess: humanoid robots in warehouses within 3-5 years. In factories within 5-7 years. In homes within 10-15 years. Each stage requires a different level of reliability, safety, and cost reduction.

The warehouse timeline is the one I’m most confident about. Controlled environments, repetitive tasks, high economic incentive, and the investors who just put $2.6 billion into Figure are the same companies that operate those warehouses.

Why humanoid specifically

The shape matters. I wrote about this before, but it’s worth repeating: the world was built for humans. Doorknobs, stairs, shelves, tools, vehicles, all designed for a body with two arms, two legs, hands with fingers, eyes at roughly 5-6 feet off the ground.

A robot that fits the human form factor can operate in any environment built for humans without modification. No special ramps. No custom shelving. No redesigned tools. It just works where humans work.

That’s the pitch. That’s why the money is flowing to humanoid form factors instead of specialized robots. A humanoid robot is a general-purpose machine. You don’t build a new factory around it. You put it in the factory you already have.

The feeling

I grew up on Asimov. His robots were humanoid. They were part of society. They had rules (the Three Laws) and jobs and relationships with the humans around them. It was fiction. Beautiful, speculative fiction.

$2.6 billion from Amazon, Microsoft, and NVIDIA is not fiction. It’s a bet. A large, calculated, strategic bet that Asimov’s vision is about to become operational.

I’m not saying we’ll have positronic brains and Three Laws compliance anytime soon. I’m saying the physical shell, the humanoid body that can work alongside humans, is no longer a dream. It’s a funded program with a timeline and the backing of the biggest technology companies on Earth.

The robot economy is coming. The checks have been written.

Related thinking: