Correlation after the fact could work. Arrest a few hundred drug traffickers over the course of several years, then feed their plate numbers and the past decade of everyone’s traffic data to the AI.
Your and my threshold for acceptable risk seems to be similar. But we’re taking a risk. If you graph that risk, it lowers predictably based on time cooked, and the rate at which it lowers increases at higher temperatures. That was my only point, and I’m not sure if you’re disagreeing or just providing anecdotes and your preferences.
Safe is a spectrum, I suppose. I eat rare beef and runny eggs, but there’s always a safety warning at the bottom of the menu. Still, the level of “safe” is a function of both temperature and time, at least according to the USDA.
That apostrophe moves faster than William Lane Craig’s goalposts.