Err... definitely not "regional slang." Literally the dictionary definition of the word "hatch" i.e. "to emerge from an egg."

So, welcome to (checks notes) 1967, I guess?

I'm Pro-nuclear power, but I do have to agree that "containment leak" is definitely a very understated way of saying "you can see the pretty blue light from kilometers away."

I don't believe today's models are AGI, but "not understanding cost" is a dangerous assumption to make. If you read the METR analysis of the Hugging Face attack, the agents used were very much aware of their own compute budgets and worked to maximize what the swarm had available to achieve the task.

Yeah, Schumer of all people whining about the surveillance state is absurd. Voted for the PATRIOT Act in 2001. Railed against Snowden a decade later. But, sure, Chuck, we totally believe you're on the side of the American People here, rather than simply being smart enough to realize that supporting Flock is a fast track to being voted out of office in the current climate (and a bipartisan one at that).

Qualified Immunity is a red herring. Don't get me wrong, it's a whole cloth creation of the Supreme Court that flies in the face of section 1983 and Congress should have fixed this problem decades ago, but QI is, first of all, limited to civil rather than criminal liability and, secondly, every police force in the country indemnifies its officers in the event of civil litigation that arises out of their actions.

If you want a real fix here, it's holding police accountable. Not throwing up hands and crying that "the big bad police union said that if we arrest officers for murder they'll go on strike, so what can we do?" Not only should their jobs not be protected when they commit obvious wrongdoing, they should be held to a higher standard in terms of criminal liability than Joe Citizen is in general. (Conversely, bona fide good faith error should be treated as just that).

Tell me you didn't read the fucking article without telling me you didn't read the fucking article.

No, Flock did not indicate a 100% match, the alleged 'detective' queried for "all dodge durangos" in a given area and chose one that was a different color than the vehicle reported in the accident... which had no damage on it. This is, 100%, on the cop and not Flock.

That's not the issue - insurance companies have no problem giving insurance to humans which have "individual learning capabilities."

The problem is that while insurance companies can extract compensation from humans - essentially they can coerce the humans to work "for them" to pay them back - they can't get compensation from the AI, and there isn't a good framework to get compensation from the AI developers.

The issue isn't that the AI learns - the issue is that our legal framework doesn't have a good way to get "compensation" from errant AI - because the AI itself doesn't care if it gets shut off. You can't hold the AI itself "liable" like you can a person. So until we change the liability framework, we'll be stuck. This isn't a technical problem, so much as a social one.

Part of self-driving cars being safer than people is that they don't drive "very close together" - they follow at the appropriate following distance and speed for the lines of sight. This alone accounts for a large portion of the improved safety performance of these systems.

Autonomous cars are actually very good at the driving task. Where they fail is when they encounter situations that are "abnormal" and I think the industry has a problem here; state of the art is either to train specifically for every situation (impossible in the limit) or to kick the decision to a human in a support center somewhere. Things like emergency vehicles, school buses, farmer markets, construction zones, etc.

The industry needs to figure out how to get their systems to "get confused, then reason their way out of a situation."

This long tail is where humans still do better - when there is some "new" scenario, like a particular combination of debris in the roadway, on a Tuesday, with gawkers, during a fireworks display. You will never be able to "train" your way out of that - you need the correct structural formulation for the problem space.

The big takeaway I have from their study is that their study areas - major metropolitan centers - is that cities have way, way worse than average incidents per vehicle mile travelled.

I mean the national average is only something like 1.25 fatalities per 100M miles, injuries only 75 per 100M miles, and "all" incidents like 140 per 100M miles. But the benchmark numbers on the linked websites range from 200 to almost 700 injury incidents per 100M miles in those cities.

Like, if you are using the average number, you can't even reduce the number of incidents by 841 over 271 million miles, because the expected value is only 203. You have to have a rate of at least 300 injury-incidents per 100 million miles to reduce by 841 at over 271 million miles. So really, their results are driven by San Francisco, Austin, and Atlanta.

Their confidence bands are also pretty low; for example they quote 2 serious or greater injuries in San Fran, which on 82 million miles there is a rate of "0.02 per million miles". But 2 injuries is not really enough to make a statistical inference.

Least of all because all the miles are with different versions of their software - you can't make any predictive claim.

Statistics never lie, but you can make them say whatever you want...

I just wanted to let you know that you have successfully driven me off of /., after some 25 years.

Not just you, obviously, the site has been in steep decline for a long time, but you were such a huge asshole that you managed to be the last straw. So, congrats for that. I won't be back.

More like 99.99%. Maybe need some more nines in there. Large C codebases without regular critical CVEs are basically nonexistent.

What bankruptcy-prevention actions are they asking for?

15-25 kWh drawn by the utility is quite a bit! I hope they're paying him well for that.