this post was submitted on 08 Jul 2025
29 points (96.8% liked)

Fuck AI

3376 readers
626 users here now

"We did it, Patrick! We made a technological breakthrough!"

A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.

founded 1 year ago
MODERATORS
 

I've seen some sites/programs mentioned before for this, but thought it may be good to have a solid list.

What tools do we have today to identify slop, whether it's video, audio, text etc.? I know right now most of us can identify it just because it's off or feels wrong in some way, but we are going to need better tools in the future to be able to truly tell. Also bonus if it's an offline tool.

all 18 comments
sorted by: hot top controversial new old
[–] Cocopanda@lemmy.world 3 points 7 hours ago (2 children)

I’m dyslexic, have ADHD and basically can’t read. Is it terrible that I use Ai to make me seem more capable when I need to write things?

[–] ZDL@lazysoci.al 1 points 5 hours ago

No. There are legitimate use cases for LLMbeciles, and as long as you are wary of hallucination yours is a perfectly cromulent case. Another case is, again providing you check for delusion, translation. LLMs are far superior to whatever tech Google Translate, Baidu Translate, et al use when performing translation. As long as you double-check that you didn't get hit by an LLM hallucination.

[–] truxnell@aussie.zone 1 points 5 hours ago

Not at all, there are valid uses for LLMs and I take this community with a gain of salt. I think a lot of us passionately hate the hype narrative and useless uses for the tech, but it doesn't mean that every single use of it is inherently bad.

The biggest thing if worry about might be the risk of it giving/outputting misleading information if it's easy for you to sense check it.

[–] ZDL@lazysoci.al 2 points 9 hours ago

Develop a taste only for art whose origin can be definitively proved to predate 2000. Eschew all art whose creation date has four digits and starts with a 2.

[–] 6nk06@sh.itjust.works 7 points 19 hours ago

Disdain and contempt for its users.

[–] wise_pancake@lemmy.ca 7 points 20 hours ago (2 children)

As soon as you have a public tool to do this, bad actors will use that tool to fine tune models and evade it.

[–] ZDL@lazysoci.al 1 points 8 hours ago

I'd love to see the tool used to fine tune a model based on "read, look at, and watch only stuff made pre-2000".

It'd be one Hell of a tool that incorporates time travel!

[–] hoch@lemmy.world 2 points 15 hours ago (1 children)
[–] Kolanaki@pawb.social 5 points 13 hours ago (1 children)
[–] a_person@lemmy.world 6 points 13 hours ago

Rule 34 is all ai at this point its unstandable

[–] Saledovil@sh.itjust.works 2 points 15 hours ago

Critical thinking mostly.

[–] Blaster_M@lemmy.world 3 points 18 hours ago
[–] cronenthal@discuss.tchncs.de 4 points 19 hours ago

I believe the market will sort this out. AI slop will only stay relevant in the domain of scammers and spammers, where demand doesn't count.

[–] gratux@lemmy.blahaj.zone 3 points 20 hours ago

I don't know of a tool to reliably identify AI works. Usually I resort to a vibe check, and if I'm not sure, I check the rest of the profile.

[–] NoneOfUrBusiness@fedia.io 1 points 17 hours ago (2 children)

What do you mean "defend" against slop? How do you defend against a social media post?

[–] ordinarylove@lemmy.blahaj.zone 3 points 8 hours ago

🙄 "what do you mean defend from harmful content"

[–] bridgeenjoyer@sh.itjust.works 1 points 13 hours ago

Was thinking more along the lines of detecting Ai "fingerprinting" in images and video.