The two history about AI and art. The new history and no so new. And joy and furious. And furious joy. And may be a little funny.
Fetch error
Hmmm there seems to be a problem fetching this series right now. Last successful fetch was on February 06, 2023 06:07 ()
What now? This series will be checked again in the next day. If you believe it should be working, please verify the publisher's feed link below is valid and includes actual episode links. You can contact support to request the feed be immediately fetched.
Manage episode 345711021 series 3389184
The first history.
People Can't Stop Feeding Their Selfies Into a Super Mean AI.
ROASTERBOT 5000.
There's a hot new AI on the block, but instead of generating images, this one analyzes them and spits out crude roasts of anyone they depict.
The AI, known as the CLIP Interrogator and created by a generative artist who goes by the handle Pharmapsychotic, is technically a tool to figure out "what a good prompt might be to create new images like an existing one."
In reality, though, it tends to spit out descriptions of people that can be mundane, puzzling, staggeringly insensitive and, at times, admittedly a bit hilarious.
It told one user she looked "tired and drunk," for instance, and accused another of having a "deformed face." It told a Georgia Tech professor that he looked like a "muppet" with a "punchable face." It called another a "weak beta male."
When this blog's writer inputted their own photo, in fact, they were presented with the artist Lizzo's name, some porn-ish terminology, and for some reason a tag for the 2019 Ari Aster film "Midsommar" (which, it should be noted, is one of their favorite movies).
DETAILS HAZY.
It's tough to say what the mean-spirited AI is meant to accomplish. We know it's built on OpenAI's Contrastive Language-Image Pre-Training (CLIP) neural network that was released in 2021, and hosted by Hugging Face, which has dedicated some extra resources to deal with the crush of traffic.
But beyond that, it's pretty mysterious: is the creator trying to give us insight into how AI peers back at us, or did they tune it to be as mean as possible, or is something else entirely going on?
Needless to say, we're fascinated. Futurism has reached out to Pharmapsychotic for comment about their AI generator, but they haven't gotten back to us yet.
Until they do, we'll all be left wondering why, exactly, this neural network is so incredibly mean — though if it remains as over-trafficked as it is right now, a lot of people will be spared from its harsh tongue.
Thanks for the histories from our friends of the futurism.com.
81 episodes