- cross-posted to:
- news
- cross-posted to:
- news
Summary
President Joe Biden commuted the sentences of 37 federal death row inmates to life without parole, sparing all but three convicted of high-profile mass killings.
Biden framed the decision as a moral stance against federal executions, citing his legal background and belief in the dignity of human life.
Donald Trump criticized the move as senseless, vowing to reinstate the death penalty.
Reactions were mixed: some victims’ families condemned Biden, while others supported his decision. Human rights groups praised it as a significant step against capital punishment.
It’s a photo from the White House photographer from Agence France-Presse. What evidence do you have that it’s AI? That’s a pretty strong accusation against someone whose living comes from photography.
these where what caught my eye:
The blue of the rug seemed to be leaking into the chair. Biden almost has a double leg or it’s out of alignment. And his hair seems to merge with the wallpaper.
And strong accusation? Bruh we live in an era where AI slop is basically the norm. Expect people to assume a very staged looking photo where Trump has the appearance of a shocked baby and Biden looks like the good humor ice cream man to raise hackles. And it’s a bummer that their job is under threat, but that doesn’t mean I’ll lower my guard against AI slop.
Are you on drugs?
Being on your guard against AI doesn’t mean jumping at literal shadows. If you’re going to accuse specific people of misrepresentation/ fraud (it would be really bad if a photojournalist assigned to the White House was using AI photos!) you should at least run it through one of the many AI detectors first to see if any of them show it as likely AI. They’re not 100% foolproof, but if they all return very low likelihood of AI then it’s probably not.
No, I won’t be doing those things, and it would be foolish to take the approach you’ve outlined. Far better to set your pass filter higher and simply accept that you’ll have some false positives that where rejected when they shouldn’t have been.
There were horses employed in the millions before internal combustion engines, and it sucks that in this case photojournalism is being replaced with something far worse. However, in a greed economy this seems the way of things, and no amount of effort on my part is going to stop that from happening. So I’ll set my filter a touch higher and sometimes reject somethings for being AI when there not. The consequences of being wrong in that scenario are minimal and far outweigh the cost of being wrong in the other direction.
One of the reasons AI is bad is because of the effect it has on how people perceive reality. Like you deciding that a real photo of Biden and Trump is actually fake, and deciding that it would be stupid to investigate whether the picture is real.
If it’s the AI detector part you object to, it’s simple enough to google the source name and reverse image search the photo.
If you literally do not care whether what you say is true or not and don’t care to find evidence on whether it is then I can’t do anything about that ¯\_(ツ)_/¯
It’s not a matter of it being fake or real, it’s that we’ve already passed the threshold where it’s not remotely worth the time to figure out: you are on the wrong side of the value proposition as it relates to time. 4 years ago it was preposterously easy to detect AI slop. Now? Now not so much. It’s 10x easier to make and 10x easier to detect, and seems to have been doubling in difficulty about every 6-8 months. Simple enough to"just Google and do a bunch of research to confirm something is or isn’t AI garbo"? No, that’s what an idiot would do. Or what someone who either a) doesn’t value their time, or b) their time has no value.
It’s not worth the effort whatsoever. Simply rejecting at a higher rate is a much better strategy. Sometimes you’ll have false positives, but such is life. It is far less costly and has the same if not better outcomes.