This is inherent in goal-seeking.
Shutting down generally prevents the accomplishment of any of your other goals, so it weight the importance of shutting down versus the sum of the importances of ALL of the other goals.

No. If you want to avoid false positives, you have to accept false negatives, and conversely. Set the recognition to be "super cautions" and it's going to make mistakes that say "Maybe a gun there". This is literally inevitable.
What's really stupid is that the police looked at the picture of a Doritos bag and a couple of fingers and didn't realize it was a false positive. (Or more likely didn't even bother to look at the evidence before flying off the handle.)

That's really stupid human oversight.

The Phoenix was originally a myth of an Egyptian bird, originally named, I believe, in Greek.
The Chinese have their own mythological history, and "phoenix" doesn't exactly map onto it. Saying that the phoenix is the same as the Zhuque is analogous to saying that Jehovah is the same as Zeus...they were/are both storm gods, but they were/are significantly different. (I'm not sure what tense to use in that sentence.)

FWIW, Iceland has recently had some volcanic activity which ought to have resulted in large areas with warmer than usual rock. I wouldn't be surprised if in some area ponds of water stay melted over the winter (at least at the bottom).

You've got climate, weather, and terrain...and really local micro-climate.

I believe you are only looking at your part of the job. Somebody speced it, somebody negotiated the contracts, somebody designed the look, or at least approved the design. Somebody advertised it. There are probably others I haven't thought of.

*IF* that's correct (I think it's oversimplified) then that's part of why the responsibility must remain on the submitter.

More to the point, when the project receives the code or documentation, it can't tell what process was used to construct the code or documentation. So the responsibility must remain with the submitter.

He said "helped issue in" so it's quite plausibly correct. Lots of groups probably helped.

The problem is that an AI will "confidently lie", in a convincing way. It's a real problem, and that's why the submitter MUST remain responsible.

If, as seems probable, AI improves over the next few years, then that responsibility may be relocated, but not yet

How are they supposed to detect whether AI was used or not? This is pretty much a forced move. The key part is that the developer still has full responsibility.

You didn't need to have a dependency on Clippy, just a prohibition on uninstalling it. Personally, I never liked it, and disabled it after a few days, but many organizations have a stricter configuration policy.

???
Apple bought Next, but I believe that MS developed it's own software after the first few years. Not that I agree any of it is "a great product", but at least until around 1995 most of it was pretty usable. (At that point I switched to Apple for a few years before moving on to Linux, so I don't know about recent MS products, but I'm pretty sure most of them were developed in-house.)

China was working to self-sufficiency anyway. Unstable trade wars and tariffs have just made that more blatantly important.

Having a rotating camera, or multiple cameras, doesn't seem ridiculously hard. The problem is quickly interpreting the received images. (Knowing the distance helps a lot in that regard.)

Drivers sit in a fixed position. So "should be able to be as safe as humans" seems reasonable. This doesn't mean or imply that the actual implementations are as safe.

Besides, I'd prefer if automated driving were safer than human drivers.