Every few months, another company unveils a humanoid robot. Two legs, two arms, a head. It walks across a stage. The crowd applauds. And then nothing happens — because humanoid form isn't first-principles design for anything except replacing a human body, which is only useful in environments already built for humans where humans can't go. Like Mars.
Form follows function. Humanoids follow neither.
Engineering works by deriving form from function. You need to fly? Build wings. You need to dig? Build a drill. You need to cut? Build a blade. The human body is brilliant, but it's optimized for a very specific set of evolutionary pressures: climbing trees, running across savannas, storing energy as fat, surviving predators, having babies. None of these are relevant to helping you learn piano or cook dinner. A humanoid robot in your kitchen is a body designed for the African grasslands 200,000 years ago, repurposed for a task it was never optimized for. That's not engineering. That's cosplay.
The uncanny valley is a feature, not a bug
Humans read form instinctively. When something looks human, we expect it to behave human. We judge it by human standards. And we feel — on a deep, preverbal level — that it's competing with us. A humanoid robot in your house triggers the same instincts as a stranger in your house. Is it going to take over? Is it watching me? Is it going to replace me? These aren't rational fears. They're evolutionary ones, and no amount of friendly marketing will override them. You don't want something that looks like it could kidnap your kids or steal your partner standing in your living room. You want something that feels safe, warm, and clearly not trying to be you.
Humans expect function from form
When you see a chair, you know you can sit on it. When you see a knife, you know it cuts. Form communicates function. A humanoid communicates: I do what humans do. Which means people expect it to do everything a human can — walk, talk, reason, emote, improvise. When it inevitably falls short (and it will, for decades), the disappointment is crushing. The form overpromised and the function underdelivered. Muse doesn't have this problem. Its form communicates exactly what it does: it sits on your table, it moves expressively, it projects light, it watches and listens. There's no gap between what it looks like and what it is.
The right form for the right job
If you're building a robot to navigate a human-built space station, humanoid form makes sense. The doorways are human-sized. The tools are human-shaped. The controls are designed for human hands. But your home isn't a space station. Your home is a place where you cook, play music, draw, do yoga, help your kids with homework. The right robot for these activities doesn't need legs, or arms, or a torso. It needs eyes that see what you're doing. A voice that guides you. A projector that puts information where you need it. And a presence that feels like a companion, not a competitor.
Muse's form is its function
Muse is designed from first principles for one job: being the best possible companion for real-world activities. Its rounded body feels approachable, not threatening. Its movement is expressive but clearly non-human — you read it as personality, not as imitation. Its projector turns any surface into an interactive workspace. And because it doesn't look like a person, you never feel like it's competing with you, judging you, or trying to replace anyone in your life. It's a tool that feels like a friend. That's hard to build. And it starts with not building a human.