2 ms·

I guess there's two ways to read a ratio like that. Either A) the false positive rate rounds to 100.00000%, or B) the correct positive rate rounds to 100.00000%

by graypegg 10mo ago

I guess there's two ways to read a ratio like that. Either A) the false positive rate rounds to 100.00000%, or B) the correct positive rate rounds to 100.00000% and the few that slip through are great tragedies due to "just not investing enough", thus making the false positives worth it. I'm glad B) at least wasn't made /explicit/ in the article, but damn... they do point at it by implication. You're totally right about the juxtaposition being manipulative. > The company isn’t aware of any school shootings where its tech was deployed. A thing that happens 50 times a year, across the entire US has not occurred at any of the small number of pilot schools... where apparently "threats" occur multiple times a day? EDIT: confirming I agree, emphasis on /explicit/.

dragonwriter 10mo ago
> I guess there's two ways to read a ratio like that. Either A) the false positive rate rounds to 100.00000%, or B) the correct positive rate rounds to 100.00000% and the few that slip through are great tragedies due to "just not investing enough", thus making the false positives worth it. I think for (B) to be a justifiable reading, the national stats would have had to have been much higher before the roll out, with a significant share of those national stats being from the particular schools that happened to be the leading implementors. But, yeah, I agree that that is a possible implication of the presentation on the surface.