@briankrebs@infosec.exchange @campuscodi@mastodon.social This is just going to lead to alert fatigue, and people tuning out alerts.
It already happens with lane assist. I rented a car for a 6-hour drive through the desert (flat, straight, dull) and lane assist was good enough that I didn't have to correct it... except for the panic-inducing alarm every 15 minutes telling me to put my hands on the wheel. The first time, I almost got rear-ended due to sensory overload - Everything started beeping and flashing, so I slammed on the brakes while my brain tried to figure out what threat I needed to avoid (and slamming on the brakes when traffic is averaging 80-90mph all around you is a bad idea). Eventually I saw the "put hands on the wheel" warning. WTF, toyota.
The thing was, my hands were on the wheel. I just didn't need to correct it, so the car couldn't tell... so it would flash lights and alarming beeps constantly. I eventually learned to just ignore those beeps, and I'd just tug at the wheel randomly to avoid them.
Thinking about an AI camera trying to figure out WTF is happening and blaring alerts all the time sounds like a great way to distract drivers into MORE accidents, not less. It's entirely predictable, but in 15 years we'll see investigations and committees trying to discover what went wrong.
Meanwhile, car companies selling adverse information so insurance companies can increase premiums will net tons of money for certain people, and other people will die due to distractions while driving. Guess which end of the economic spectrum will receive the better end of the deal...