this post was submitted on 15 Nov 2025
85 points (97.8% liked)

Hacker News

3027 readers
310 users here now

Posts from the RSS Feed of HackerNews.

The feed sometimes contains ads and posts that have been removed by the mod team at HN.

founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
[–] Carnelian@lemmy.world 3 points 5 days ago

No I know what you mean, it just doesn’t track

Even if it's, say, 80% accurate under optimal conditions

Why are we using 80% as an example when it’s definitely impossible for the most skilled interpreters to beat a coin flip? Right like you see the issue? The problem isn’t that it’s sometimes unreliable the problem is that is fundamentally impossible to derive any meaningful information from it

What you’re arguing for is similar to the thought experiment where they have an airport scanner that can tell with 99% accuracy whether someone is a terrorist (spoiler: even with this level of accuracy, the scanner cannot be relied on alone, because counterintuitively due to false positive rate, the end result ends up being that only one in a thousand positive scans is a true positive).

But the polygraph is simply not similar to that thought experiment. It’s like “narrowing your search results to better manage your resources” based on the results of asking a magic 8 ball. Which is why it’s illegal to submit polygraph as evidence in court