KI kann aus leeren Gesichtern politische Orientierungen vorhersagen – und Forscher befürchten „ernsthafte“ Datenschutzprobleme
https://www.foxnews.com/politics/ai-can-predict-political-orientations-blank-faces-researchers-fear-serious-privacy-challenges
8 Kommentare
„Researchers are warning that facial recognition technologies are „more threatening than previously thought“ and pose „serious challenges to privacy“ after a study found that artificial intelligence can be successful in predicting a person’s political orientation based on images of expressionless faces.
„I think that people don’t realize how much they expose by simply putting a picture out there,“ said Kosinski, an associate professor of organizational behavior at Stanford University’s Graduate School of Business.
„We know that people’s sexual orientation, political orientation, religious views should be protected. It used to be different. In the past, you could enter anybody’s Facebook account and see, for example, their political views, the likes, the pages they follow. But many years ago, Facebook closed this because it was clear for policymakers and Facebook and journalists that it is just not acceptable. It’s too dangerous,“ he continued.“
„But you can still go to Facebook and see anybody’s picture. This person never met you, they never allowed you to look at a picture, they would never share their political orientation … and yet, Facebook shows you their picture, and what our study shows is that this is essentially to some extent the equivalent to just telling you what their political orientation is,“ Kosinski added.
The authors concluded that „even crude estimates of people’s character traits can significantly improve the efficiency of online mass persuasion campaigns“ and that „scholars, the public, and policymakers should take notice and consider tightening policies regulating the recording and processing of facial images.“
According to the point map, Republicans appear to be fatter than Democrats (or R simply requires a little higher number of close-up headshots).
I read some of the underlying study, and to me it looks like the AI was able to identify political orientation as well as humans, but that both the AI and the humans weren’t very good at it. The prediction correlations were fairly weak.
Genie is our of the bottle, I think.
Unless we go Covid mask or religious garb, anyway.
If you did male=repub, female=dem, you’d be above chance already so not very impressive.
While it may be trying to do that… I wonder how far off it would be automatically guessing left-leaning for certain non-whites, and women? In the current political climate, while it sounds extreme, it’s also not unreasonable.
This also gets into the utility of the info, while I can know the political affiliation of whomever is looking at my display/whatever, what exactly could I do with it? You’re not going to see an ultra-right wing candidate win over college students in any meaningful way. Outside of one specific candidate, you can’t get away with changing your political messages/stances for each person you meet.
It’s a cool idea, predictive elements, but I suspect it’s not going to be able to be meaningful in the longrun.
No mention of how the sausage is made? At all? Come on now I’m genuinely curious.
…you know what though? I could see this be very useful in movie casting. I’m like 90© sure that the Sabrina sequel used AI to dig up the right people to cast cause of just how eerily close to form the core cast were. Now imagine using a broad concept punched into the database to choose who would instinctively feel right for the portrayal of a given idea.
Cause I’m pretty sure whatever the machine did humans probably do to a degree as well subconsciously.
I can look at some white people from Arkansas and figure they have a good chance to be conservative.
Does that mean I am an AI?