Hacker News Reader

Comment by kibwen | Hacker News Reader

No, neither the hardware nor software is anywhere even remotely ready, where by "ready" we mean "safe to share living spaces with unsupervised children and pets without EVER accidentally reducing your toddler to a fine paste, literally a 0% chance". That's the minimum that people will accept, and that's more than ten years away, if it ever happens.

View on HN

Both of the preceding comments are true. These things are about to arrive in vast volume, because the factories to build them are already starting to run. And they're nowhere near ready for that.

It took over a decade for Waymo to get from "able to drive around SF for demos" to "3x safer than humans, with thousands of vehicles on the roads".

A lot of these things may end up in closets, next to the VR headset and the 3D TV.

Meanwhile, there's a tired mom of triplets who's wondering "how fine? (a paste)"

You'd be surprised. I think Waymo have already proven this; not perfectly safe, but below the care threshold. And the demand for childcare is huge. Of course, what then happens is how the ensuing child neglect case falls out.

We're probably going to end up with the situation where the burden "it is considered criminal child endangerment to leave your child alone with the robot" falls on the parents, not on the robot manufacturers.

That's not how product liability laws work in the U.S. or in the E.U.

The first time a child gets harmed by a robot, the company making that robot will go spectacularly bankrupt. If a child gets killed by a robot, it would likely end the consumer robotics industry for a decade or more.

Teslas have killed probably hundreds of people at this point and the only changes to self-driving laws have been to allow more of it.

Tesla sales have cratered in all but 2 countries.

People tried to hold gun manufacturers liable for dead children, and the law was simply changed to exempt them.

Robots and AI have too much money to be held liable.

Perhaps, but also the US has the 2nd amendment and it seems (from the outside) to be a big part of their culture.

I don't think AI in general has that benefit despite people occasionally trying to criticise AI companies who restrict the output of LLMs by reference to the 1st amendment; robotics in particular definitely doesn't have that benefit.

Good news, the current apocalyptic memory prices will mean that only apocalyptically rich people will be able to afford them at the toddler-pasting stage if they come out any time soon!

Reply on news.ycombinator.com