BARBIE (2023) Dir. Greta Gerwig +letterboxd Reviews By Men

BARBIE (2023) dir. Greta Gerwig +⭐letterboxd reviews by men

Baby Onesie Designs

Galaxies on Ice

A reader wrote in a while ago with a suggestion: they were about to have a baby and wondered if I could use AI to come up with some new ideas for baby onesies. I can't find the letter any more, and I don't remember how long ago it was. But I finally figured out how to make it work! Reader, if your kid has already outgrown the onesie section, I believe these can be sized up to toddler.

Dandelion Underwater

Two things made this work:

I used BLOOM rather than chatgpt or the other largest language models. The big language models are too good at copying existing designs from their training data.

I asked it to generate its list as if it was a page on AiWeirdness.com

Fart Whale

Ketchup metaphors

In my opinion, the most interesting creative use of large language models is to generate text that's nothing like a human would have written. If your AI is just going to lift human creative output virtually verbatim, you're not only shortchanging the humans you could have hired to write similar things, but also plagiarizing the original humans from the training data. In that sense, BLOOM, with its less-perfect retrieval of human output, is better at this task than GPT-4.

It is creepy to me however that the only reason this method gets BLOOM to generate weird designs is because I spent years seeding internet training data with lists of weird AI-generated text.

Pumpkins on Mars

Onions in Snow

Meatball launch controls

Turn them into real life cursed artifacts and put them on a small child who doesn't know what they're wearing

More (and rejected designs) at aiweirdness.com

100 posts!

barbenheimer x pinkfloyd :3

reminder for the twitter, reddit and tiktok migrants: please do not censor words here. if you censor them, people’s blacklists and mute functions wont work, especially for important content warnings (this goes for twitter as well altho that function is breaking over there). spell out the whole word in the post and the tags. do not use euphemism words like unalive. the algorithm here does not work that way (there is an algorithm. not everyone uses it, everyone uses the following tab).

the only time you should censor words is when you do NOT want them to show up in the tag, like if you’re saying something unkind about a ship you should not tag it or mention directly the ship name so it does not clog the tag, or if you’re mentioning a person or community who should not get attention or clout, or whose attention you do not want to attract, then censor it (the way one might on twitter to avoid term searching)