Even if this fails only in 0.1% of cases, given enough users this will kill babies.
(Of course you should compare this to humans, but in any case, do the math! And get a good lawyer.)
(Of course you should compare this to humans, but in any case, do the math! And get a good lawyer.)
It is amazing/horrifying to me how many people are intent on reincenting the reasons why we have UL, the FDA, the FCC, traffic laws, seatbelts, electrical codes, fire marshals and unions.
It's not going to solve all problems.
Parent 2: Nah, the baby monitor would have warned us.
In other words, if you want AI to take over tasks of humans you either do it well or not at all.
Tesla autopilot could steer you into a concrete wall, a baby monitor with some LLM attached to it does not directly harm the child.
This is stupid
(I just can't eat a whole one.)