My kids would ask me, “What is the movie about?”, and I’d tell them, “About 90 minutes.”
That’s an old Dad joke, and I don’t miss it.
I mean, there’s not really any good textual answer to that question. The real world answer is to sit the kid in a car and say “Like this”.
So who wants to provide the three- or four-year-old we can use to test how well one can drive compared with an AI?
Would you settle for eight?
I would except none of the AIs are up to the sophistication, intelligence and common sense of the typical eight-year-old.
And I think it’s a great illustration of “live by the sword,” (misinformation is free speech, there can’t be any standards for fear of who’d enforce them), “die by the sword” (howd’ya like this misinformation, still feel the same about non-enforcement?)
I 100% agree it should be broadcast on every available social media system. The more the better.
Because, they are not “AIs” any more than a Commodore-64 was an “AI”. The programming tech is quite sophisticated, but it is desperately lacking. The most critical area is when something like a fully-automated system encounters an edge- or corner-case that has not been covered in its training: the typical response for the model is to improvise or to decide that the situation is not important, which results in things like vehicles running into overturned semi trailers because they cannot resolve what that thing is, or chatbots just making shit up because an answer is expected.
The current generation of “AI” is nothing more than a glorified computer, but fiendishly harder to debug.
I like to describe them as “a chatbot with a search engine stapled to it”.
“Chatbot” meaning “ELIZA” and “search engine” meaning Lycos circa 1996.
You know, it’s been awhile since the media has tried to blame violent crime on video games. It almost feels nostalgic at this point.
Which isn’t that different from a human brain…
Other than the amount of stuff that is a “corner case”.
I personally have never had to avoid running into an overturned semi. I’m not even 100% sure I’ve ever seen one in person, as opposed to pics of them. OTOH, I have great confidence I can recognize one, and given enough space and time for my slow human cognition, realize it is an obstacle to be avoided.
These AIs may be able to handle (made up number) 99.9999% of driving situations. But with (WAG) 200 million drivers doing wacky shit out there every day, 99.9999% still leaves a very large amount of possible shit un-covered. So AIs driving surprisingly badly will continue to be an everyday occurrence.
They’re working towards 99.9999999999%, but it’ll take awhile to get there.
Among us? Wow - talk about reaching.
That basic game premise/mechanic has existed in playground games forever; I’ve played it as Werewolf and Mafia. My son plays it today as “Imposter” with his friends at school. Ridiculous to blame it for contributing to a crime.
All future AIs will also be glorified computers. Kind of in the definition.
Hell, we’re just glorified computers.
If someone is implying that this game turned someone into a real life assassin…
…then someone is clearly desperate for a new spin on this story.
I think there’s maybe a bit of irony, but no more than someone who likes to play tag who is killed in a high speed chase. Doesn’t seem to merit an actual news story.
Oh yeah - it definitely smelt of Slow News Day.
This is a Tesla on autopilot. Clearly, if it fails to see the obstacle in time to slow down, it is out-driving the headlights – something which human drivers do, but they ought not.
