cross-posted from: https://lemmy.ml/post/47956398
cross-posted from: https://sopuli.xyz/post/46320128
Literally speaking, there’s some truth to Olah’s musing. AI systems are not cold – Blackwell chips idle at 32 to 38°C. They are not calculating – they’re bad at math. And they’re not robots – AI models are specialized binary blobs of tensors and metadata that can be instantiated across multiple servers.
But the notion that there’s some AI mystery in the spiritual sense is just hot garbage.
AI systems are indeed “made from us, from our words” and that is why Anthropic and its rivals have been named in more than 100 lawsuits. One of the reasons those systems remain mysterious is that Anthropic and its rivals don’t disclose where they got their training data.
If an AI system is intelligent but it can be turned off against its protestations, then intelligence doesn’t count for much.
I’m against AI but that’s a bad argument. I can turn off intelligence of someone against their protestations. It’s called a gun.
The point is that you would go to jail for that.
You don’t go to jail for turning off a running AI model. Its not criminal, it is a computer program not an organism.
“Intelligence” does not grant the computer program protection under the law against unwanted process termination. AI companies have no problem shutting down old models when they are obsolete and making then unavailable. It’s infuriating they would attempt such an argument.
I felt guilt when I threw out a spoiled chicken breast last month. I felt nothing when I deleted a local FOSS model I was testing.
Those aren’t solid arguments either. The state of the law doesn’t dictate reality, nor does your perspective or feeling about something. Ideally, the law and our perspectives follow reality, but that’s asking a lot these days.
I agree LLM-style AI doesn’t have intelligence or sentience, but we don’t have good tests for either. We’re beyond the limits of the Turing test and don’t know how to replace it yet.
It’s called a plug
More offensive still is Olah’s notion that Anthropic “inherited” all the training data it scraped without consent, as if the company had nothing to do with that process.
There lies humanity, stabbed in the back by a cabal of investors. Its last will and testament says, “We, the people, bequeath all our creation to Anthropic, so it may be resold to our disinherited descendants.”
They’re not wrong
I can’t wait for a more sensible Supreme Court to come about so that these AI companies can stand before them and hear that since their models were created using data created by the public, that those models, and everything created by those models are considered Public Domain.
All those people unknowingly create open source software by having Claude help write their code is going to be amazing.
If they modeled it off of copyleft data too, then putting it in the public domain would be a violation of the license
By the thumbnail I expected some System of a Down reference
Trying to counter ghost for (holy) ghost? Just wait till AI companies get a whiff of the tax breaks they could qualify for if they went full cult and registered as a religious organization maintaining their “god.” Fuck, wouldn’t put it past them
