Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

fossforce.com

Post reply on HN
Print view View original

21–30 of 99 posts

Re: Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

#21

Post by thesuperbigfrog »

visarga wrote: post #4

> “As I see it, ‘intelligence’ needs something like the ability to know or understand some area,” he said. “If something can’t actually understand things we shouldn’t say it’s intelligent, not even a little intelligent, but people are using the term artificial intelligence for bullshit generators.” I wouldn't go as far in blaming AI. It is certainly error prone and has very short autonomy. It's not better than subjec…

>> I thought that as a pioneer in the free software movement, he would fully support LLMs that democratize skills and make them universally applicable.

RMS worked on early AI research at the MIT Artificial Intelligence lab in the 1970s.

LLMs are fundamentally different than many of those early AI systems in both goals and how they work.

It is likely that he views them through a different lens and sees their shortcomings in sharp relief compared to the early AI systems he worked on.

Re: Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

#23

Post by steve1977 »

mik1998 wrote: post #20
munchler wrote: post #13

[flagged]

A statistical model cannot understand anything since it does not think in the first place.

That implies that thinking is not based on some kind of biological statistical system.

Re: Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

#24

Post by vidarh »

ad404b8a372f2b9 wrote: post #14
vidarh wrote: post #10

Earlier quoted context omitted.

His thinking here leans on assuming that there's a qualitative difference between LLM's "understanding" and our "understanding" that justifies using the term for us, but not for LLMs. Maybe there is, but their reasoning ability is sufficient that I don't think is at all a given, and at a minimum an assertion that should make you ask someone how they define "understanding", because defining that in a way that is focus…

I don't know that his argument is based on essentialism since he goes on to describe algorithms which he considers artificially intelligent: "There are programs that can look at a photo of some magnified cells and tell you, with greater likelihood of being right than any human doctor, whether it’s cancerous or not,”. And he puts it in contrast to his previous statement about LLMs not being artifically intelligent. It…

I think he's seriously contradicting himself there in that the algorithms he argue are AI certainly do not meet his stated standard, and it seems like he's drawing a line where he refuses to use the term about anything that might be seen as intelligent that doesn't meet his standard for intelligence, but is fine with using it for techniques where there's no chance of confusion.

I think if he wasn't making an essentialist argument about LLMs he worded himself exceedingly poorly. Arguing against the danger of blindly believing LLMs is good, but at the same time he's destroying his own credibility on the subject by exaggerating how bad they are vs. focusing on the dangers of blind trust.

This is especially notable when he then goes on to point to measurable success as a metric for other AI tech - he'd have made a far better point if he argued that LLMs also need to be evaluated carefully on a case by case basis like these other systems, rather than be taken on trust.

Re: Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

#25

Post by steve1977 »

munchler wrote: post #13

[flagged]

Not so quick with disqualifications…

Stallman was working at the MIT Artificial Intelligence Lab, long before most of us have first heard the term AI or were even born.

Also, for example: „While working (starting in 1975) as a research assistant at MIT under Gerry Sussman, Stallman published a paper (with Sussman) in 1977 on an AI truth maintenance system, called dependency-directed backtracking.“

Source: https://en.m.wikipedia.org/wiki/Richard_Stallman

Re: Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

#26

Post by simiones »

Interesting to see he is very clearly against what IBM Red Hat is doing. Not surprising, since it was the obvious thing to say about it (they are trying to prevent you from using the full rights given to you by the license). Still, nice to see no ifs, ands or buts: what they are doing is ethically wrong and should stop. It may or may not be illegal as well, but it's certainly not a good thing.

Re: Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

#27

Post by tannhaeuser »

munchler wrote: post #13

[flagged]

RMS while studying at MIT published about what's considered early/classic "AI" (for example I was very surprised and double-checked the author when I encountered a paper about Prolog backtracking / backjumping by him many years ago). I'd assume he knows the topic he is writing about very well.

Re: Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

#28

Post by simiones »

vidarh wrote: post #24
ad404b8a372f2b9 wrote: post #14

Earlier quoted context omitted.

I don't know that his argument is based on essentialism since he goes on to describe algorithms which he considers artificially intelligent: "There are programs that can look at a photo of some magnified cells and tell you, with greater likelihood of being right than any human doctor, whether it’s cancerous or not,”. And he puts it in contrast to his previous statement about LLMs not being artifically intelligent. It…

I think he's seriously contradicting himself there in that the algorithms he argue are AI certainly do not meet his stated standard, and it seems like he's drawing a line where he refuses to use the term about anything that might be seen as intelligent that doesn't meet his standard for intelligence, but is fine with using it for techniques where there's no chance of confusion. I think if he wasn't making an essentia…

> This is especially notable when he then goes on to point to measurable success as a metric for other AI tech - he'd have made a far better point if he argued that LLMs also need to be evaluated carefully on a case by case basis like these other systems, rather than be taken on trust.

The examples he gives are technologies where you can point the AI to a problem and trust the results will work as intended. You don't need to check every individual output of those AIs, you can just trust that it will be better than the output you'd get from a human expert (even though it's of course not 100%, and a combination of AI + human expert may be better).

What is the equivalent for LLMs?

Re: Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

#29

Post by boomboomsubban »

visarga wrote: post #4

> “As I see it, ‘intelligence’ needs something like the ability to know or understand some area,” he said. “If something can’t actually understand things we shouldn’t say it’s intelligent, not even a little intelligent, but people are using the term artificial intelligence for bullshit generators.” I wouldn't go as far in blaming AI. It is certainly error prone and has very short autonomy. It's not better than subjec…

His objection seems predominantly pedantic, and bemoans that "antisocial media" is profiting off them. That is incredibly in character for Stallman, he doesn't seem opposed to them as a whole.

Re: Richard Stallman Talks Red Hat, AI and Ethical Software Licenses at GNU Birthday

#30

Post by oddevan »

steve1977 wrote: post #23
mik1998 wrote: post #20

Earlier quoted context omitted.

A statistical model cannot understand anything since it does not think in the first place.

That implies that thinking is not based on some kind of biological statistical system.

Counterpoint: person hears 15 arguments. 14 say one thing, 1 says another. Person agrees with the 1 argument because it resonates with them more.

I'm not saying the 14 are right and the 1 is wrong. I'm saying that the decision was not based on a statistical model.

Post reply on HN