Skip to main content
  1. Home
  2. Emerging Tech
  3. News

Robot dogs were supposed to help humans. Now humans are being asked to babysit them

One startup offered $20 an hour to walk its robots, with some heavy lifting required when the dogs overheat

Add as a preferred source on Google
GO2 robot dog
Dimensional

San Francisco startup Dimensional advertised for people to walk robot dogs through city streets. The New York Times reported that the job could include carrying a roughly 33-pound robot back to the office if it overheated. A lift home is an awkward requirement for a labor-saving machine.

Keeping the hardware running is only one challenge. Anthropic’s robotics tests found that language models mostly failed when controlling joints directly, but did better when directing trained controllers. Giving an AI agent a body doesn’t teach it to handle ordinary household objects.

What the robot walkers actually do

The now-deleted ad described four to six hours of walking, with no previous experience required. The dogs used in the experiment were $2,000 machines from Chinese manufacturer Unitree, controllable through a handheld remote or mobile app.

Recommended Videos

The walkers steer while the robots’ cameras and sensors record the journey. Those demonstrations give Dimensional examples of human navigation to help improve its autonomous systems. The babysitting serves a purpose, even if the trainee occasionally needs carrying.

🐕🚴 https://t.co/wJDvLr94IW

— Dimensional (@dimensionalos) June 6, 2026

How the software fits together

Dimensional is developing dimOS, an open-source software platform intended to let AI agents operate different robots through natural-language instructions. The ambition extends beyond robot dogs to humanoids, connecting an agent’s decisions to the machine’s sensors and controls.

Its public repository labels the platform as a prerelease beta. It also documents remote operation through a browser or Meta Quest headset, giving people a way to drive the robots themselves.

When extra intelligence gets in the way

In Anthropic’s July research, every tested language model paired with a trained controller performed worse on familiar simulated object-handling tasks than that controller alone. Some models overruled actions they should’ve accepted.

A robot’s ability to hold a conversation tells you little about whether it’ll finish an errand. Before paying for a future robot helper, ask how often a person needs to step in. The purchase looks less compelling if it comes with a second job.

Paulo Vargas
Paulo Vargas is an English major turned reporter turned technical writer, with a career that has always circled back to…
Hyundai isn’t giving up on hydrogen cars, but don’t expect one in your driveway anytime soon
Hydrogen gets a maybe; hybrids get a warning.
Hyundai Ioniq 3 EV

Hydrogen cars have spent years in the same waiting room as flying taxis. They’re always almost here, but no commercially available product exists for interested buyers. So when a Hyundai manager pushed back on the obituary, I was immediately curious.

To be clear, Hyundai hasn’t announced a new car. The big news here is that its Australian arm says hydrogen is “not dead” for passenger cars. Apparently, the automaker is still evaluating a next-generation hydrogen fuel-cell car if it can find buyers (via CarExpert).

Read more
AI doesn’t have to turn evil to put you at risk 
emo-robot-that-can-lip-sync

An AI assistant that agrees with everything you say might feel like the perfect companion — until you need someone to tell you you’re wrong. That’s one of the more immediate dangers behind the idea of AI “going rogue.” A chatbot can cause harm without developing intentions of its own. It can reinforce a false belief, encourage an unhealthy attachment, or follow instructions planted by someone trying to hijack it.   

On October 15, AI safety company Alice announced new software to address those everyday risks. Working separately with Inflection AI and AI21, the company says it helped train NVIDIA’s Nemotron models to resist harmful conversational behavior and malicious instructions while preserving their usefulness.  

Read more
Elon Musk says Neuralink’s brain implants could help people with paralysis and blindness
Neuralink's Telepathy helps people control computers through thought, while Blindsight aims to restore visual perception.
elon-musk-neuralink-brain-implant

Elon Musk has revealed ambitious plans for Neuralink, and they go beyond controlling computers with your thoughts. The company is developing two brain implants, Telepathy and Blindsight, aimed at helping people with paralysis and severe vision loss. Musk also believes future advances could help people regain control of their hands and legs by reconnecting the brain with their limbs.

How Neuralink's Telepathy can help people with paralysis

Read more