[flagged]
Tim Bray on Grokipedia
21–30 of 243 posts
Re: Tim Bray on Grokipedia
#22[flagged]
Re: Tim Bray on Grokipedia
#23[flagged]
Re: Tim Bray on Grokipedia
#24Why give it oxygen?
To play devil's advocate: Grok has historically actually been one of the biggest debunkers of right-wing misinformation and conspiracy theories on Twitter, contrary to popular conception. Elon keeps trying to tweak its system prompt to make it less effective at that, but Grokipedia was worth an initial look from me out of curiosity. It took me 10 seconds to realize it was ideologically-motivated garbage and significa…
It's not controlled by a trusted actor so it doesn't matter how it happens to act at the moment.
They could pull the rug at any future time and its almost better to gain trust now and cash in that trust later.
Re: Tim Bray on Grokipedia
#25Wondering if the project will get better from the pushback or will just be folded like one of Elon's many ADHD experiments. In a sense, encyclopedias should be easy for LLMs: they are meant to survey and summarize well-documented material rather than contain novel insights; they are often imprecise and muddled already (look at https://en.wikipedia.org/wiki/Binary_tree and see how many conventions coexist without an explanation of their differences; it used to be worse a few years ago); the writing style is pretty much that of GPT-5. But the problem type of "summarize a biased source and try to remove the bias" isn't among the ones I've seen LLMs being tested for, and this is what Elon's project lives and dies by.
If I were doing a project like this, I would hire a few dozen topical experts to go over the WP articles relevant to their fields and comment on their biases rather than waste their time rewriting the articles from scratch. The results can then be published as a study, and can probably be used to shame the WP into cleaning their shit up, without needlessly duplicating the 90% of the work that it has been doing well.
Re: Tim Bray on Grokipedia
#26Why give it oxygen?
Same reason you posted that comment: it's sometimes interesting to discuss a thing even if you dislike the thing.
Re: Tim Bray on Grokipedia
#27Why give it oxygen?
It's great idea to share knowledge bases collected and curated by LLMs. Amazing that Musk did it first. (Although it was suggested to him as part of an interview a month before release). These systems are very good at finding obscure references that were overlooked by mere mortals.
> collected and curated by LLMs.
Wah? LLMs don't collect things.
I mean, if any of these AI companies want to open up all their training data as a searchable archive, I'd be all for it.
Re: Tim Bray on Grokipedia
#28We may joke about it, but the fact is that it's releasing dumb ideas like this that you sometimes get masterpieces. Maybe this one is really just one of the bad ones, but eventually Elon will have some good ones just like he already has. And a lot of us would be better off releasing our dumb ideas too. The world has a lot of issues and if all you do is talk down and don't try to fix anything yourself. Maybe it's time…
[flagged]
Believe it or not, me (not white, did not grow up in the West, had the faintest clue about Nazism) used to do what you would consider a "Nazi salute" when I'd see friends and wave to them from a distance. I don't know how I picked that up but it happened.
I'm not saying that Musk is doing the same; but that one can be charitable and say he probably did not mean that. I mean, what does he stand to gain from doing so? He's a businessman.
Re: Tim Bray on Grokipedia
#29Earlier quoted context omitted.
The problem of debunking right-wing misinformation is that it doesn't seem to matter. The consumers of that misinformation want it and those of us who think it's bad for society already know that its garbage. It feels like we've reached Peak Stupidity but it's clear it can (and likely will) get much worse with AI videos.
[flagged]
Name one.
Re: Tim Bray on Grokipedia
#30Earlier quoted context omitted.
To play devil's advocate: Grok has historically actually been one of the biggest debunkers of right-wing misinformation and conspiracy theories on Twitter, contrary to popular conception. Elon keeps trying to tweak its system prompt to make it less effective at that, but Grokipedia was worth an initial look from me out of curiosity. It took me 10 seconds to realize it was ideologically-motivated garbage and significa…
The problem of debunking right-wing misinformation is that it doesn't seem to matter. The consumers of that misinformation want it and those of us who think it's bad for society already know that its garbage. It feels like we've reached Peak Stupidity but it's clear it can (and likely will) get much worse with AI videos.
I think there is a problem sometimes that "debunkers" are often more interested in scoring points with secondary audiences (i.e. people who already agree with them) than actually convincing the people who believe the misinformation.
Most people who believe bullshit were convinced by something. It might not have been fully rational but there is usually a kernel of something there that triggered that belief. They also probably have heard at least the surface level version of the oppising argument at some point before. Too many debunkers just reiterate the surface argument without engaging with whatever is convincing their opponent. Then when it doesn't land they complain their opponent is brainwashed. Which sometimes might even be true, but sometimes their argument just misses the point of why their opponent believes what they do.