people are sleeping on the last part of his comment:
"the best thing to me, throughout all the history of capitalism or innovation or whatever you want is to just flood the market"
JDAM all the data centers. right now. **urgently**
Cory Doctorow, Chris Hayes, and David Roberts: why do people hate AI i can't figure it out
the AI companies: we are going to take away your ability to think and sell it back to you, that's our plan, literally and explicitly, we are saying it from a stage.
He's using LLM-based spellcheckers, and kind of implying this is new amazing technology.
Presumably to prove that he isn't a "luddite" or whatever.
@jonne @thomasfuchs @malcircuit @peter
If I have been reading his complaints correctly Cory never hated AI unless it was burdening the user: i.e. "Reverse centaur"
I guess that's good enough to sell a lot of books because it is a "functionally smart" position with a little pushback against the AI trend, and not enough people in the media are willing to give us even that, but -my- problems with AI go deeper than the interface. I still consider it (attempted) intellectual theft and a sanitized interface for environmental destruction that provides results which are a lot less helpful than all the stuff that came before AI -- all the stuff they deliberately took away to make AI seem useful
(I recently stopped letting Cory in my feed after he started to share substack articles)
@RnDanger @jonne @thomasfuchs @peter
I feel like it's important to keep in mind the context of the article. He's using a spellchecker LLM. It's not a chat bot. He's not asking it questions. He's not asking it to write for him. It's like criticizing someone for using a keyboard app with autocorrect.
I also have a deeper philosophical opposition to most uses of LLMs, but a spell checker is such a trivial application that I'm having a hard time thinking of a reason it's "bad".
@malcircuit @jonne @thomasfuchs @peter
He's running olama on a local computer. I haven't looked into the the training for that model but if i was able to be convinced that there's an ethical model to use it might be that one.
So he's set up an AI at home. That's not "Big LLM", but it's also definitely not "anti AI", which a lot of his fans are
@RnDanger @jonne @thomasfuchs @peter
Yeah, I know nothing of Ollama, so it's hard for me to take a position on it. At the same time though, previous generations of spell checking and grammar checking are also arguably "AI" technology, and trained on similar sorts of datasets. I'm not sure whether a spellchecker based on lower complexity neural nets and Markov chains is meaningfully different from one that's based on an LLM. Seems more like just a matter of scale.
@malcircuit @jonne @thomasfuchs @peter
Well one difference is that he sends his whole work through at once and says "find errors" instead of looking for squiggles in the text as he goes, which sounds good to me because those squiggles distract me so much from actual writing.
Another is the training. Where's the data from? Did those people agree to provide it? How much energy did it take to train? I just don't know these things.
@RnDanger fwiw many word processors have the option to do a full spellcheck on the whole text (without any LLMs); indeed that was the default until the 90s.
I agree that this is way less distracting and a byletter workflow.
Anyway, he should just hire a good editor—who for example could tell him not to insult his audience…
Peter
Jonne Arjoranta
Mallory's Musings & Mischief
Human Brain Enthusiast
2xfo