News for hackers. Stuff that matters.

A comment on Mistral's Robostral Navigate: a state of the art robotics navigation model. Parent comment

Re:Mistral's Robostral Navigate: a state of the art robotics... by philipkglass on Wednesday July 08, 2026 @11:44AM (#48833449)
It seems like a stronger story for robotics, since smaller models can always react to the environment faster than large models for a given level of hardware resources. Also because robots that keep their models local for latency or robustness reasons aren't going to be carrying many kilowatts of inference capacity.
[ Parent | on HN ]
Re:Mistral's Robostral Navigate: a state of the art robotics... by lumost on Wednesday July 08, 2026 @12:04PM (#48833686)
remote inference should be sufficient for most robotics applications with potentially a small model for safety critical actions running locally.

Unless you are in military robotics or automotive of course :)

[ Parent | on HN ]
Re:Mistral's Robostral Navigate: a state of the art robotics... by AlotOfReading on Wednesday July 08, 2026 @12:13PM (#48833789)
There are many, many factories that still don't have internet access on the floor, and commercial inference generally has response latencies measured in seconds. I struggle to imagine a factory spending hundreds of thousands for the local compute to run a large model either, given how cheap they are about expenses.

I'm also skeptical that you can cleanly differentiate between "safety critical actions" and "actions", though this is less of a practical concern given how laissez-faire some manufacturers are. For context, I work on safety critical robotics (in automotive).

[ Parent | on HN ]