AI with a body in 2035: five honest predictions
Not science fiction, not hype. Five guesses about what robots with AI brains will (and won't) be doing in about ten years, from a small studio that builds software for a living.
By Duangduean Studios
Predictions about the future are usually wrong in interesting ways. In the 1960s, people expected flying cars and moon holidays by 2000, and nobody predicted that everyone would carry a computer in their pocket.
So take these with a pinch of salt. They're our honest guesses, based on where the technology is today, about what AI with a physical body will look like around 2035.
1. Most robots won't look human
Humanoids get the headlines, but most useful robots will be shaped for their job: wheels in warehouses, arms in kitchens, small tracked machines on farms, drones for inspection. A human shape makes sense where the space was designed for humans and the tasks keep changing, like a car factory or a home. Everywhere else, simpler shapes will win because they're cheaper and more reliable.
2. Factories and warehouses first, homes much later
Today's plans already point this way. Hyundai is talking about deploying more than 25,000 Atlas robots in its factories, starting in 2028. By 2035 we'd expect humanoids to be fairly ordinary in big factories and logistics centres.
Homes are harder: every house is different, there are children and pets, and the robot has to be cheap enough for families. Our guess is that by 2035, home robots will exist in wealthier homes much like early dishwashers did: useful, expensive, and still a bit clumsy.
3. Your phone's assistant will "drive" many machines
Today you talk to an AI assistant on your phone. In ten years, we think that same assistant will be able to operate machines around you: tell the robot vacuum to do the kitchen, ask the delivery robot to leave the parcel at the back, check on an elderly parent through a home helper. The "brain" will often live in the cloud or on your phone, and the robot body will be more like a tool it uses.
4. New jobs will appear, and some old ones will change a lot
Every big technology shift destroys some jobs and creates others. Heavy lifting, repetitive assembly and dangerous inspection work will increasingly go to machines. At the same time, there will be new work: robot technicians, trainers who teach robots new tasks, safety inspectors, remote operators and people who design how robots and humans work together.
The honest worry is the transition. People in jobs that shrink don't automatically become robot technicians. How well countries handle retraining will matter far more than the robots themselves.
5. Trust will matter more than technology
The robots that succeed won't necessarily be the smartest. They'll be the ones people trust: clear about when a human is watching, safe around children, honest about their limits, and easy to switch off. We think privacy and safety rules for home robots will be one of the big political debates of the early 2030s, the way social media rules were in the 2020s.
What we're doing about it
As a small studio, we're not building robots. But we build software and games, and everything above runs on software. Simulations where robots learn, apps that control them, games that let people try new ideas safely: this is the kind of work that gets more important as AI gets a body.
Ten years from now we'll look back at this list and laugh at what we got wrong. That's fine. The point of predictions isn't to be right. It's to start thinking early.
Further reading: Hyundai Motor Group plans deployment of 25,000 Atlas humanoid robots (Humanoid Guide) · Tesla Optimus robot launch timeline (TechRepublic)