3 ms·
I wonder if you could ever make an ML model that could give you a reasonable estimate of the risk of taking an action would be. "If I'm in a low-crime neighbor
by dekhn 4y ago
I wonder if you could ever make an ML model that could give you a reasonable estimate of the risk of taking an action would be.
"If I'm in a low-crime neighborhood on a normal temperature day, what is the risk that leaving my children in a car for 10 minutes with a window cracked and the doors locked will lead to my children being injured, kidnapped, or killed?"
The risk isn't nonzero (kids left in cars do, at some frequency, get injured, kidnapped, or killed) but it's probably far smaller than other risks people accept (such as driving the car around with the kids in it, or walking tired kids through the parking lot while also pushing a cart).
It would be great if we had such things, because it seems like people don't actually have good judgement for risks (understatement). My guess is huge parts of our day involve taking actions we think make us safer, but in fact do not.
- tepitoperrito 4y agoWouldn't this be something actuaries have a database for? Or is that a stretch... EDIT: I nerd-sniped myself and am sketching out how I'd design the schema for those tables : /
- moistly 4y agoOne presumes they did not walk to the car, to go sit in it; no, they got in the car, travelled some distance, and then were left in the car. The act of driving them is many orders of magnitude more dangerous than the risk of sitting briefly in the car. Driving is by far the most dangerous thing you regularly do. All else pales by comparison.