I’m not sure where the hell I heard that certain body fluids are sterile (probably from ammonia or something?), but that makes such little sense. It’s literally your wastewater.

I’ve unironically untangled my thought process sometimes in the midst of trying to write a useful prompt (something more explicit than “it no worky”). The text box is more helpful than the model like 50% of the time.

Huh… I should build one of those sometime 🤔

My client is glitching out on the gif and making it extra funny

Aaahhh, CWM… Better, more exciting times… Now we’ve got locked bootloaders and system integrity mechanisms everywhere 🙄

I never knew this until somebody else told me because I almost can’t smell anything

Most international market phones are supposed to do GLONASS just fine though? How does that work?

The joke is a cross between a famous anecdote about Diogenes and the inner workings of vision-capable AI models.

Diogenes really liked to disrespect his social superiors, so when, in a lecture, Plato defined human beings as “featherless bipeds”, Diogenes plucked a chicken and presented it to him saying “Behold - Plato’s human being” (or something along those lines, the quote is slightly different in every source).

In order to “see” things from images and video, a garden variety LLM needs to integrate with another model that analyzes the picture it is given and gives a text description (source: HuggingFace). What I’m getting at is that, based on that idea of stripping things of context, ChatGPT could probably confidently tell you something in the vein of Plato taking the piss due to its inherently limited ability to see.

Behold, a man! - ChatGPT

a wonderful picture? u mean a list of colors and a columns count?

Isn’t that all digital photography anyway?

Many front cameras have variable focus, which I believe has the same effect. Software does most of the heavy lifting in phone photography, so there’s also a lot of correction being applied at all times anyway.