Photographer and musician. Former journo at VICE, Playboy, NYT and more. Sausalito, CA.
Sausalito, CA
Joined October 2011
- Tweets13K
- Following2.7K
- Followers636
- Likes6.6K
incredible that a *single US agent* was able to penetrate Venezuelan security forces and convince Maduro and his wife to board a NY bound flight just by holding up a picture of a cat and saying the words "6 - 7" whenever questioned
if "AI" just means admitting the symbolic was never real, and Sam is our final webmaster, then fair. And I wonder what kind of vegetables he will elect to grow, how he will depict their fruit, how he will prepare his harvest into supper, and amongst whom he SHALL DINE
Even on a 'healthy diet', LLMs rot as a function of their own scale and social adoption, since their outputs will be ingested to train future models' inputs.
So, now is the golden age of synthetic media, and for the same reasons 2006-2016 was the golden age of organic media :)
This might be the most disturbing AI paper of 2025 ☠️
Scientists just proved that large language models can literally rot their own brains the same way humans get brain rot from scrolling junk content online.
They fed models months of viral Twitter data short, high-engagement posts and watched their cognition collapse:
- Reasoning fell by 23%
- Long-context memory dropped 30%
- Personality tests showed spikes in narcissism & psychopathy
And get this even after retraining on clean, high-quality data, the damage didn’t fully heal.
The representational “rot” persisted.
It’s not just bad data → bad output.
It’s bad data → permanent cognitive drift.
The AI equivalent of doomscrolling is real. And it’s already happening.
Full study: llm-brain-rot. github. io
my upcoming title " Electronic Textuality Under Twilight of Meaning Collapse " was accidentally deleted when i tried to 'select all' then hit two other buttons inside a one-level-of-undo framework, and will not be available in time for the holiday season
overhearing someone explain their new app / tech idea in public in 2025 literally sounds like someone describing a trap door in a lemonade stand
hello world
tech bro 1.0- nerds in a garage who want to change the world, burning man, psychedelics, Silicon Valley, super technical
tech bro 2.0- 996, blue chip pedigree from FAANG or university, gym bro, obsessed with tracking everything, AI vibecoding marketing slopmaster
tech bro 3.0- philosopher, well-read, highly sociable, ultimate craftsman, unattainably put together, high hospitality, high taste
we are obviously still in number 2, but I feel like the liberal arts renaissance man type will rise to prominence in the tech world
One Battle After Another blindly fulfills the dystopian prophecy of the book it's based on, which is kind of a bummer, man. letterboxd.com/dearwolf/film…
My 🥹 review of One Battle After Another. It's sad to see Pynchon get vibejacked into culture war ammo. I loved Vineland so much, and PTA did such an incredible job adapting Inherent Vice, but this ain't that - dis one for the illiterati 🥂 cheers guys boxd.it/bbphgn
hey @apple for iOS 26 pls make "add to reading list" in Safari one click instead of two (now under share click) and also make it so my screenshots just save to photos with one click/swipe (now under ✔️ click)
Practitioners are almost always left out of "task reductions" in the AI discourse. Most people who do anything will look at the way technologists caricature their practice and say, "that's not what I do at all".
This from a well established copywriter: substack.com/home/post/p-171…
GPT is like the guy at the party who takes a super insightful little thing you say and paraphrases it 10x to the entire room, pedantically formalizing all the obvious grammatical relations, and trumpeting in some zippy new terms from a champagne-stained thesaurus.
i think one of the great services humanists can do is find fail cases that people can try themselves to "feel the edge of the system" ways that people can come to a more visceral and heuristic stance towards the LLMs