«The Wikimedia Foundation, the nonprofit organization which hosts and develops Wikipedia, has paused an experiment that showed users AI-generated summaries at the top of articles after an overwhelmingly negative reaction from the Wikipedia editors community.»
It's nice to see places where community can help course correct!
https://www.404media.co/wikipedia-pauses-ai-generated-summaries-after-editor-backlash/
@gedankenstuecke On the one hand yes, on the other hand it's not promising that it's just been 'paused'
@joepie91 as I said elsewhere, I give them the benefit of the doubt for having taken the step to re-evaluate, which hopefully can also include 'fully stopping'.
@gedankenstuecke @joepie91 I just wonder who is pushing for this. Call me conspirational, but isn't it abundantly clear that the present generation of AI is not up to this?
@martinvermeer @joepie91 I think the problem they are trying to address is very real. A lot of articles in WP are written by experts for experts, having some more accessible summaries would be great. And it's not an easy problem to solve (but as you say, AI summaries won't solve this one either)
@gedankenstuecke @martinvermeer What really gets me here is that this is indeed a problem that people have been complaining about for years, but it's also a problem that has been functionally *ignored* for years - with attempts at rewriting an article to be more accessible, for example, being reverted because it is "not in an encyclopedic tone".
And it makes me very much question what the real intent is here, and whether the accessibility isn't more of an excuse to play with a new tech toy.
@joepie91 @martinvermeer fair enough, though I would guess that the people deciding to try those generated summaries aren't the same as the problematic volunteer editors that make the reverts
@gedankenstuecke @martinvermeer Quite possibly, but if they essentially need a planet-boiling labour exploitation machine to bypass internal governance problems instead of just addressing the editor behaviour and policies, then that kind of feels like all hope is lost already
@nichtich @gedankenstuecke @martinvermeer I have exactly zero interest in AI booster hype.
There exist zero working models that solve the problems with LLMs/GenAI and there is no reason to believe this will ever change due to fundamental properties of the technology, and that is where the discussion ends for me.
@joepie91 @gedankenstuecke @martinvermeer that's a perfect example of being dogmatic, but that's ok. In doubt, I'd also first say no any current application of AI.
@nichtich @gedankenstuecke @martinvermeer Dogmatism is the remaining option when you're dealing with a billion-dollar industry hell-bent on exploiting and wrecking all of society
@joepie91 @gedankenstuecke @martinvermeer there a models with much less requirements and I bet there are more reasonable use cases with Wikipedia. The problem is not AI but when and how to make use of it.