As someone who works with AI _all_ the time, I recently took @delafina777.bsky.social 's survey on this and scored exactly 30%
Outside of very limited conditions it is incredibly difficult to tell when it's done right, and odds are you have already seen AI content and not known it.
Outside of very limited conditions it is incredibly difficult to tell when it's done right, and odds are you have already seen AI content and not known it.
I wish more people knew that there’s virtually no way to prove media or text is AI-generated unless you created it yourself. “AI detectors” aren’t proof. Among other glaring issues, this makes it more difficult for journalists to report on media being AI-generated even when it seems obvious
October 16, 2025 at 6:07 AM
As someone who works with AI _all_ the time, I recently took @delafina777.bsky.social 's survey on this and scored exactly 30%
Outside of very limited conditions it is incredibly difficult to tell when it's done right, and odds are you have already seen AI content and not known it.
Outside of very limited conditions it is incredibly difficult to tell when it's done right, and odds are you have already seen AI content and not known it.
“Unauthorized AI garbage done without the permission of Kaleb or his loved ones”
(Hat-tip to @delafina777.bsky.social)
(Hat-tip to @delafina777.bsky.social)
Please report this fucking AI slop. Goddamnit. www.amazon.com/KALEB-HORTON...
October 1, 2025 at 10:02 PM
“Unauthorized AI garbage done without the permission of Kaleb or his loved ones”
(Hat-tip to @delafina777.bsky.social)
(Hat-tip to @delafina777.bsky.social)
#Delafina777-AI
I have yet to be impressed with anything I've seen LLMs write.
(strap in, this is going to be a bit of a discussion)
I have yet to be impressed with anything I've seen LLMs write.
(strap in, this is going to be a bit of a discussion)
December 27, 2024 at 11:48 PM
#Delafina777-AI
I have yet to be impressed with anything I've seen LLMs write.
(strap in, this is going to be a bit of a discussion)
I have yet to be impressed with anything I've seen LLMs write.
(strap in, this is going to be a bit of a discussion)
#Delafina777-AI
the language machines are already not good at language and are not getting better at it
the language machines are already not good at language and are not getting better at it
December 22, 2024 at 2:09 AM
#Delafina777-AI
the language machines are already not good at language and are not getting better at it
the language machines are already not good at language and are not getting better at it
(7/12/23) #Delafina777-AI
More of the internet than we initially thought is already AI-generated, and that's incredibly short-sighted.
LLMs seem like they can write moderately well, right now, because they’re drawing from a high-quality body of human-created work.
More of the internet than we initially thought is already AI-generated, and that's incredibly short-sighted.
LLMs seem like they can write moderately well, right now, because they’re drawing from a high-quality body of human-created work.
December 21, 2024 at 4:54 AM
(7/12/23) #Delafina777-AI
More of the internet than we initially thought is already AI-generated, and that's incredibly short-sighted.
LLMs seem like they can write moderately well, right now, because they’re drawing from a high-quality body of human-created work.
More of the internet than we initially thought is already AI-generated, and that's incredibly short-sighted.
LLMs seem like they can write moderately well, right now, because they’re drawing from a high-quality body of human-created work.
December 21, 2024 at 3:26 AM
#Delafina777-gender #Delafina777-AI (AI companions)
December 21, 2024 at 2:26 AM
#Delafina777-gender #Delafina777-AI (AI companions)
(5/21/23) #Delafina777-politics #Delafina777-AI
“It’s legal to use AI that steals from artists so saying it’s bad is classist/ableist” is a helluva take.
You get that just as visual gen AI is trained on art w/o artists’ permission, gen voice AI is trained the same way on voices?
Legal != ethical
“It’s legal to use AI that steals from artists so saying it’s bad is classist/ableist” is a helluva take.
You get that just as visual gen AI is trained on art w/o artists’ permission, gen voice AI is trained the same way on voices?
Legal != ethical
December 21, 2024 at 12:43 AM
(5/21/23) #Delafina777-politics #Delafina777-AI
“It’s legal to use AI that steals from artists so saying it’s bad is classist/ableist” is a helluva take.
You get that just as visual gen AI is trained on art w/o artists’ permission, gen voice AI is trained the same way on voices?
Legal != ethical
“It’s legal to use AI that steals from artists so saying it’s bad is classist/ableist” is a helluva take.
You get that just as visual gen AI is trained on art w/o artists’ permission, gen voice AI is trained the same way on voices?
Legal != ethical
November 2024 is a tough month for many reasons
But AI getting the @delafina777.bsky.social treatment may be exactly what you need to keep your chin up
But AI getting the @delafina777.bsky.social treatment may be exactly what you need to keep your chin up
And as we're seeing, that's enormously powerful! You can do all kinds of very cool, almost magical (or almost *human*)-seeming things with that prediction ability.
But as one of my engineer friends is fond of saying, you can statistics syntax but you can't statistics semantics.
But as one of my engineer friends is fond of saying, you can statistics syntax but you can't statistics semantics.
November 26, 2024 at 6:08 PM
November 2024 is a tough month for many reasons
But AI getting the @delafina777.bsky.social treatment may be exactly what you need to keep your chin up
But AI getting the @delafina777.bsky.social treatment may be exactly what you need to keep your chin up