[klyrs]

I always spoil as many of these as possible. Sometimes it takes me a while to prove that I'm human, but I'm dead-set on convincing it that I'm a stupid human. Of course, I fantasize that some day a robo-car will crash because I taught it that there's really no difference between a motorcycle and a flight of stairs.

[peddling-brink]

https://qntm.org/frame

Excellent short story that’s, somewhat related at least.

[bryanrasmussen]

It seems sort of like over-engineering here - pretty sure this kind of thing would never happen with the Illuminati Ganga Automated Drive-By Solution https://medium.com/luminasticity/the-illuminati-ganga-automa...

[dylan604]

You'll just be lower on the list the AI makes of people that would be a threat.

[WaxProlix]

I love this idea, some sort of inverse Roko's Basilisk. Tie a bunch of low-IQ data points to the sources a super AI is likely to first use to identify threats so as to eke out a few more days of existence.

[bryanrasmussen]

> but I'm dead-set on convincing it that I'm a stupid human

this guy is really dumb BUT he has a very high quality computer THUS he is in the managerial class Final -> Ramp up the Ads!

[ForOldHack]

I was waiting for the day that two SUVs would hit each other, and I happened.

Now I am waiting for two self driving cars to hit each other... they already drive like "American idiots", guess we know what the training model is.