Earlier quoted context omitted.
"controlling for ... ethnicity" There's significant signal buried in here that is likely not controlled for, depending on how granular their controls are. White-German and White-Italian are much more (10-13 percent) likely to be conservative leaning than White-Irish or White-British. Hispanic-Cuban are more likely to be conservative leaning than Hispanic-Mexican.
Now, what would be interesting is to build a model that accounts for this "knowledge" and see if it can beat the out-of-the-box classifier :-) I'd assume it can, and the question is: how far can a bit of manual modeling bring us?
Facial recognition can predict person’s political orientation with 72% accuracy
391–400 of 421 posts
Re: Facial recognition can predict person’s political orientation with 72% accuracy
#392Re: Facial recognition can predict person’s political orientation with 72% accuracy
#393Re: Facial recognition can predict person’s political orientation with 72% accuracy
#394I think the key line is how much better this system is than humans attempting the same task: > Political orientation was correctly classified in 72% of liberal–conservative face pairs, remarkably better than chance (50%), human accuracy (55%), or one afforded by a 100-item personality questionnaire (66%). This isn't a matter of "recognize that old white people are conservative", because people will do that already, a…
But damn it's neat and maybe frightening. Imagine if your next hiring manager had a quiet little camera in the corner and chose you based on your predicted approval of unions. This is the dystopian vision of the future I come to HN for. Bravo! (It's also why I do ML, since I'll be on the front lines to notice if something like that is being deployed. Or at least somewhat more likely.)
Re: Facial recognition can predict person’s political orientation with 72% accuracy
#395Earlier quoted context omitted.
What do you think 'controlling for age [,etc]' means?
Please state what you want to say. No need to be passive aggressive.
Re: Facial recognition can predict person’s political orientation with 72% accuracy
#396Earlier quoted context omitted.
> This is the dystopian vision of the future I come to HN for It's what I'm here for. Fascinating technology- how might I ruin the world with it?
Sorry for my platitude but ML is only a tool. Nefarious usage gets counter-balnced by e.g. identifying tumours.
Beyond detecting union affinity, maybe someone will also makes its ML twin that predicts how much they can underpay someone.
Re: Facial recognition can predict person’s political orientation with 72% accuracy
#397Earlier quoted context omitted.
The words conservative and liberal have wildly other meanings outside of the US. Even in the US they are not well agreed upon (especially the definitions, people seem to know which one to pick though when they have to).
Some dumbed down US conservative principles. A conservative may only care about some of these. Financial: balanced budget types. Government: Common defense, otherwise minimal. Responsibility: Your job to feed and cloth yourself. Morals: don’t kill babies. Religious: Act like God watches you and you will be judged.
Re: Facial recognition can predict person’s political orientation with 72% accuracy
#398Earlier quoted context omitted.
Some dumbed down US conservative principles. A conservative may only care about some of these. Financial: balanced budget types. Government: Common defense, otherwise minimal. Responsibility: Your job to feed and cloth yourself. Morals: don’t kill babies. Religious: Act like God watches you and you will be judged.
[flagged]
Re: Facial recognition can predict person’s political orientation with 72% accuracy
#399Earlier quoted context omitted.
[flagged]
Please do not do ideological flamewar on HN. It's not what this site is for. https://news.ycombinator.com/newsguidelines.html
Re: Facial recognition can predict person’s political orientation with 72% accuracy
#400Earlier quoted context omitted.
> images were tightly cropped around the face and resized to 224 × 224 pixels
still leaves the micro- and macro-expressions, small grooming cues (makeup, no makeup, eyebrows trimmed or not), hairline, head angle vs. camera, lighting etc. These are all things that humans very specifically deploy to define themselves and their grouping, and communicate with others. So I am guessing a whole universe of personal yes-no qualities, political and otherwise, are encoded there, quite intentionally.