Eugenio Tisselli
@motorhueso
http://motorhueso.net || http://ojovoz.net || https://mastodon.social/@motorhueso || sounds: http://cubo23.bandcamp.com || movies: http://vimeo.com/wallofsound
Two mantras for the 21st century: "Machines are not living beings" "Living beings are not machines"
AI lovers: this is the sort of fascist/idiotic slop you are helping to push forward.
That means that the bubble is just gonna swell to gargantuan proportions, and when it pops, it will be devestating. I foresee Big Tech nearly collapsing when that occurs, and people are gonna want some scalps for who caused this mess.
The AI Action Plan is even worse than I thought it'd be. A tech industry wishlist with every box ticked. It signals the further coherence of Silicon Valley and the Trump-led federal government into a cooperative, automation-bent surveillance state. It's a blueprint for dystopia.
Has AI energy efficiency improved over the last few years? Perhaps not yet in real-world deployments. My reflections on @MistralAI's recent disclosure about the environmental impact of its LLM: linkedin.com/posts/shaolei-…
Render the world as a computational structure that can be optimized. Present a system that creates the illusion of optimization. Force it into the entire world together with the threat of adapting to it or getting left behind. To make it acceptable use it to create instant "art".
Is it possible to imagine a better future where we use less computation? The effects of AI hype almost demand it. This week @danmcquillan joins @parismarx to lay out the concept of decomputing as a different way of assessing technology. Full ep: techwontsave.us/episode/286_de…
If it wasn't for this direct, aggressive presidential intervention, gen AI companies would now be legally and financially buried.
Last night, Trump sided with parasitic AI companies against millions of writers, artists, filmmakers, and musicians trying to eke out a living. The rationale is if AI developers can't steal intellectual property, then they can't feed their digital demons. Sounds good to me.
I haven't found a single peer-reviewed paper covering AI-powered productivity increase that also takes into consideration the additional work required to review, correct, and oversee AI outputs. The "increased productivity" and "socially beneficial" claims have been at the core…
"Congressional Republicans have delivered on the pro-Israel organization AIPAC’s wish list in the latest military spending bill, including tens of millions of dollars a year for the Israeli military to develop AI technologies." jacobin.com/2025/07/aipac-…
Billions of $ do not produce intelligence. This is a dumb argument. They do however create a huge incentive to pretend LLMs aren’t a dead end.
The whole idea of comparing the human brain to a computer should die in a fire.
Thinking that LLM-ish AIs could soon lead our science and philosophy and art is tied to thinking we already reached the end of science and philosophy and art anyway. Nobody thinks an LLM-ish AI trained only on variations on 17th century texts could get us to 21st century civ
very good, but a) LLMs don't exhibit a "dark forest theory of intelligence" because cognition hasn't been and won't be ever grounded in the technological paradigm that constitutes their very structure and b), we don't act like turing cops toward "AI". LLMs ARE turing cops.
Science and art are about the love of knowledge, understanding, self-expression and making meaning of our lives and the world around us.