Humanoid Industrial Designer
Coroflot, Mountain View, CA, United States
Project basis, remote, global, with potential to extend into a full-time industrial design role:
A Silicon Valley startup is looking for a designer to help define the body of a home robot people would actually want in their homes.
The “humanoid” is unusually pre-specified in culture. It is the only product category humanity imagined in detail before it existed. Most technologies emerge through iteration: people encounter a new capability and then learn what they want from it. Humanoids are the inverse. The idea stretches back to ancient Greek myths. Thanks to creators like George Lucas and Pixar, humanoids are already widely adopted, even though you will hardly find one in a home.
The reference is not technical. It is social, cinematic, and behavioral. Every child knows the spec. And every adult senses that what we are getting still is not it. Despite the sleek shells styled like electric cars, projecting seriousness and performative maturity, do you see at least one machine you would actually buy and feel comfortable keeping behind your bedroom door at night?
The more serious the design language becomes, the closer it feels to Terminator than to H.E.R.B.I.E. from Fantastic Four. The more humanoid makers optimize for “capability” in the machine-learning sense (higher accuracy, broader task coverage, more anthropomorphism, better generalization), the more unsettling the product tends to feel, drifting toward the cold competence of HAL 9000 rather than the cooperative charm of R2-D2 from Star Wars.
The problem is not whether the machine can complete the task. What matters to people is how it behaves while the task is being completed. What people evaluate is operational trust: a set of behavioral signatures humans are extremely sensitive to in anything with hands and legs.
Just watch any humanoid demo. The gap between our intuitive expectations and the actual experience is too wide to bridge. You see it in small details: the obvious hollowness when it folds laundry like a zombie, or speech that feels like a chatbot placed inside a robot body.
The primary language of the humanoid we expect is behavior: posture, balance, hesitation, yielding, small corrections, and the rhythm with which it meets the world. That is how humans read intention and competence. It is not a large language model placed inside a robot body. Instead of speaking perfect English, it might use simple sounds that are expressive but not linguistic. WALL‑E with a few sounds and his camera eyes conveys more meaning than ChatGPT can express in a thousand words.
It learns differently too. To learn to ride a bike, a conventional humanoid AI may watch thousands of videos of someone else riding one. Yet it is never actually on the bike. It is like trying to taste honey through the glass of a jar.
The humanoid we want shares our reality. It does not observe the world from a distance through models. If it learns to ride a bike, it is on the bike. It is the one who falls. It senses the world through its body as it goes: balance shifting, surfaces pushing back, small losses of control, the constant negotiation with gravity. It is subject to the same gravity, friction, fatigue, and risk that we are.
Because of that, its responses follow a logic we intuitively understand. You do not know exactly what it will do. But you know it will not walk into a wall. It will not keep moving when hurt. It will hesitate under uncertainty. It will back off under stress. And it will learn useful things simply by living alongside you.
Creature Algorithm is the architectural AI discovery that makes this possible. It gives the “home organism” a kind of “soul,” establishing the missing behavioral signatures from the start.
We are looking for someone who can help give physical form to this kind of machine — something we already recognize culturally, something strangely familiar, yet fundamentally different from everything being built today. Just look at them, cataloged at Humanoid.Guide. In essence, they are nearly two hundred variations of the same idea: machines optimized for capability in the engineering sense, with an added dose of forced anthropomorphism.
So this is not a rendering task and not another shell design exercise. The body is part of the argument.
We are looking for early concept work, not finished production design: visual direction, body studies, and a first clear sense of what a Creature-powered humanoid should feel like in the world.
This begins as a paid concept project, compensated above market average, with the possibility of extending into a full-time industrial design role if the fit is strong.
To apply, please send:
portfolio
1 reference for what Creature should feel closer to
1 reference for what Creature should avoid
CV, if useful
architect@creaturealgorithm.com
#J-18808-Ljbffr
A Silicon Valley startup is looking for a designer to help define the body of a home robot people would actually want in their homes.
The “humanoid” is unusually pre-specified in culture. It is the only product category humanity imagined in detail before it existed. Most technologies emerge through iteration: people encounter a new capability and then learn what they want from it. Humanoids are the inverse. The idea stretches back to ancient Greek myths. Thanks to creators like George Lucas and Pixar, humanoids are already widely adopted, even though you will hardly find one in a home.
The reference is not technical. It is social, cinematic, and behavioral. Every child knows the spec. And every adult senses that what we are getting still is not it. Despite the sleek shells styled like electric cars, projecting seriousness and performative maturity, do you see at least one machine you would actually buy and feel comfortable keeping behind your bedroom door at night?
The more serious the design language becomes, the closer it feels to Terminator than to H.E.R.B.I.E. from Fantastic Four. The more humanoid makers optimize for “capability” in the machine-learning sense (higher accuracy, broader task coverage, more anthropomorphism, better generalization), the more unsettling the product tends to feel, drifting toward the cold competence of HAL 9000 rather than the cooperative charm of R2-D2 from Star Wars.
The problem is not whether the machine can complete the task. What matters to people is how it behaves while the task is being completed. What people evaluate is operational trust: a set of behavioral signatures humans are extremely sensitive to in anything with hands and legs.
Just watch any humanoid demo. The gap between our intuitive expectations and the actual experience is too wide to bridge. You see it in small details: the obvious hollowness when it folds laundry like a zombie, or speech that feels like a chatbot placed inside a robot body.
The primary language of the humanoid we expect is behavior: posture, balance, hesitation, yielding, small corrections, and the rhythm with which it meets the world. That is how humans read intention and competence. It is not a large language model placed inside a robot body. Instead of speaking perfect English, it might use simple sounds that are expressive but not linguistic. WALL‑E with a few sounds and his camera eyes conveys more meaning than ChatGPT can express in a thousand words.
It learns differently too. To learn to ride a bike, a conventional humanoid AI may watch thousands of videos of someone else riding one. Yet it is never actually on the bike. It is like trying to taste honey through the glass of a jar.
The humanoid we want shares our reality. It does not observe the world from a distance through models. If it learns to ride a bike, it is on the bike. It is the one who falls. It senses the world through its body as it goes: balance shifting, surfaces pushing back, small losses of control, the constant negotiation with gravity. It is subject to the same gravity, friction, fatigue, and risk that we are.
Because of that, its responses follow a logic we intuitively understand. You do not know exactly what it will do. But you know it will not walk into a wall. It will not keep moving when hurt. It will hesitate under uncertainty. It will back off under stress. And it will learn useful things simply by living alongside you.
Creature Algorithm is the architectural AI discovery that makes this possible. It gives the “home organism” a kind of “soul,” establishing the missing behavioral signatures from the start.
We are looking for someone who can help give physical form to this kind of machine — something we already recognize culturally, something strangely familiar, yet fundamentally different from everything being built today. Just look at them, cataloged at Humanoid.Guide. In essence, they are nearly two hundred variations of the same idea: machines optimized for capability in the engineering sense, with an added dose of forced anthropomorphism.
So this is not a rendering task and not another shell design exercise. The body is part of the argument.
We are looking for early concept work, not finished production design: visual direction, body studies, and a first clear sense of what a Creature-powered humanoid should feel like in the world.
This begins as a paid concept project, compensated above market average, with the possibility of extending into a full-time industrial design role if the fit is strong.
To apply, please send:
portfolio
1 reference for what Creature should feel closer to
1 reference for what Creature should avoid
CV, if useful
architect@creaturealgorithm.com
#J-18808-Ljbffr