As usual there’s no “AI” involved at all. Just statistics on data. The nonstop grifting is further eroding the already nonexistent credibility of mainstream journalism.
It’s using genAI to make best guess portraits of ICE agents, WTF are you going on about?
ICE agents just like all law enforcement agencies, are public servants employed by taxpayers. Good luck finding any other legitimate job you can hide your identity from your employer. If you choose to serve the public, you choose to be in the public eye and should be identifiable. I could care less if they wear a mask or not, but they should be required to have some other identifiable marking if not, like a badge number, or hell, even something to identify that they are agents and not just random kidnappers would be an improvement.
As for being harassed for doing their job, if they’re doing the job in a reasonable way, they wouldn’t be harassed. Sure they wouldn’t be able to hit the quotas they’re given, but having quotas for finding criminals is a backwards concept in general and means they have to create criminals when there aren’t enough that are easy to catch. It shouldn’t be easy. It should be thorough, especially if no life is in danger from the “criminals”.
I feel like this is going to get non-combatants identified and ruin normal people’s lives.
Are you talking about ICE? Yes, they have attacked only non-combatants and ruined many “normal people’s” (wtf) lives.
No, he’s talking about you being (mis)identified as an ICE agent. Because this tech sucks.
Wait until you hear about the 200 Flock cameras you drove past on the way to work this morning which already ID’d you and reported your movements to all of the 3-letter agencies regardless of whether you were suspected of anything.
I have a bicycle.
They are already using this or worse. Theil’s Palintir is already doing tons of nasty data analytics. Add in Lucky Palmer’s drone/surveillance company Anduril, and you have the data input and attack payloads ready to go, all from Chud industries.
Don’t forget about all the info on every American the doge kids stole. They’ll be using that too.
They escalated this shit in 2025 no doubt.
Yes, this just levels the playing field so i think there is little to no downside.
Yeah this is hella sketchy. It’s MAYBE one step above saying “chat gpt, show me what this agents face looks like” and then using that to try and find a match.
Skeptical about the accuracy, given everything we’ve seen AI do so far.
The author said its 60% faulty. So everyone is an ice agent, after enough time
The author said its 60% faulty. So everyone is an ice agent, after enough time
If you read the article, it is stated that there is a team that hand reviews the matches at the end because of course, like all things, no, it is not perfectly accurate.
Yes, the “AI” grift is annoying and completely untrustable.
But as usual, there’s zero “AI” involved here.
It’s just facial recognition. Just more statistics on data.
If there’s ever a target that deserves it…
Also, I FULLY support inaccurate targets. This might be the only way fucking brainless, heartless conservatives could ever agree to regulate “AI”.
they would likely just arrest the person doing it and make it against the law to identify government enforcers.
This is a European citizen leading this project. And the EU is just as dystopian as America, just in slightly different ways. So, sure they could still just arrest someone, but probably not cooperating with America.
This one of those genie out of the bottle situation, we like some of the things AI can do yet it can do a lot of others that are not so good. Regulation won’t work as shown in this case, the work is done in Europe on a US issue. I suspect that we all will start wearing full face masks again.
“I suspect that we all will start wearing full face masks again.”
Maybe to terminally online people, but for a majority of people, masks aren’t going back up. The US population as a whole isn’t going to do significantly anything.
There’s no genie. There’s no “AI”. This is just facial recognition. Statistics run on images.
This is probably the least harmful application compared to what the imperial machine is doing.
That would be kind of sick, if masks were more normalized
It use to be said that you shouldn’t be afraid of your government however things have changed. If everyone starts wearing samurai or Halloween masks just to get some privacy it will just be another step in societal decline.
Samurai masks? This dystopian timeline is finally starting to take on some of those edgy cyberpunk aesthetics.
I am skeptical.
All this stuff works on probabilities, even with a fully visible face the model just says it’s really confident hence articles like this - Facial recognition leads to wrongful arrest.
So the more of the face that’s covered, the larger the confidence interval, or basically the less accurate it is.
Wait until you hear what empire is doing with facial recognition in palestine, etc…
Doesn’t really matter if it works, as long as it instills the fear that it could.
Where dat repo at
No “AI” exists in the making of this statistical analysis or anywhere else, but here’s a few options for facial recognition.
https://hackernoon.com/6-best-open-source-projects-for-real-time-face-recognition-vr1w34x5
Sure, let’s pretend the field of AI, with that exact name, isn’t at least 50 years old.
Facial recognition has probably been taught in AI courses for almost two decades now. “Artificial Intelligence: A Modern Approach” is 30 years old, and only about half of it is concerned with Statistics.
