I know the Internet Archive has been under a ton of infrastructural pressure lately, but anyone have any idea about how long they might take to review an application and get back to you?
-
@aud no, we don't agree with that. the world has changed in a way which makes the old solutions not work anymore. the reason that has happened is that generative models and "is this slop" models are the exact same thing, almost literally, and certainly in terms of what training data they need. whoever has more data wins... and the playing field has been leveled, so google no longer has an advantage on that front.
-
@[email protected] Right. I think to a certain extent one should avoid indexing those in the first place. Or, at least, that's how I imagine it would work.
Of course, a person could also create a bunch of reasonable sounding personal blogs, get them into the index, then change the contents to be a bunch of evil unicorns after a period of time has passed and let it slip in that way. Sigh. -
@aud yeah, you have to keep in mind that this is an adaptive system, we're not fighting against the wind or the rain, we're fighting against people who will find new forms of malicious behavior in response to any protection we come up with.
-
@aud that isn't to say we can't win, but for each proposed mitigation we need to do adversarial modeling: how would an attacker get around this? what would we do about that? how would they get around that? etc
-
@[email protected] hmmmm. yes. I'm definitely keeping my thinking too bounded by that problem, probably because I remember how good search used to be.
I think you're right that I need to free myself of it. -
@aud and mitigations that just become an endless chain of "we'll re-train the thing based on the feedback that we got it wrong at the previous step" are generally not the best idea because, like, there's no reason to think we can actually stay ahead in that arms race
-
@[email protected] ah, yeah, sorry. I really meant "solved" for a particular time and place for a particular set of cases. In the sense that yahoo, etc, had 'solved' it during their heyday, so to speak.
But perhaps that's still not correct, even with a rather loosey-goosey definition of what solving means. -
@aud oh good
sorry to belabor the point! -
@[email protected] I guess the idea of "where is the information and how do we find it?" is unsolvable so long as time keeps flowing. There's just approaches that work better than others, until they don't.
-
@[email protected] agreed.
(that's actually a big reason why we're having this discussion, in that I feel very free and un-judged to propose whatever comes to mind, whether I can see and work through the flaws or not. You're not only very insightful and helpful, but you're also very non-judgemental and safe to bounce even non-working ideas against. It's nice) -
@[email protected] Well, and then I think that comes down to a sort of fundamental issue with the approach, too, which harkens back to you saying that maybe search itself is dead. Constant 're-training' might stem the bleeding, but it also might mean your approach is just... not going to work and you're just throwing bandaids on top of it.
(see: LLMs and "fixes" for the most obvious example. But SEO spam is possibly a form of that as well) -
@aud awww!!!! that's really sweet of you to say. we try really hard at that, but we still feel like we fail a lot
-
@[email protected] No no, it's good. Even though I'm less devoted to the idea of 'search' as my language might indicate (I'm really just using it as a shorthand for 'how do get information?'), it's still very much bound up in my thinking as to how information might be obtained. I mean, the idea of asking questions and receiving an answer and judging its suitability is... well, old. As you're aware. And search engines try to present themselves as basically that in a very literal form. So as the "current existing model", they definitely dominate my brain space.
-
@[email protected] I know the feeling (as someone who tries for the same thing... most of the time. When someone is in good faith), but I can safely say I always feel that way about any discussion we've had!
-
@aud @ireneista i will note that i generally believe "relevancy" which is abstract and separate from any specific problem domain to be
(1) a direct precursor to current slop models
(2) more of a business metric analogous to "engagement"
(3) vastly inferior in all respects to technology that leverages domain expertise. when google search provided a clear set of composable rules for their distrbuted document matcher, everyone loved them! "relevance" was considered something to infer and not to demonstrate (need an even stronger verb here tbh) -
@aud @ireneista butting in with different lived experiences here but i don't at all see this as an intrinsic problem. i think this is a very poorly studied research space for various reasons but i think there is a massive amount of work to be done on public infrastructure for power and protection instead of merely austerity
-
Irenes (many)replied to d@nny "disc@" mc² last edited by
@hipsterelectron @aud (we also feel that we should point out that we've done way too little talking about library science, in this conversation so far. library scientists study questions such as how people find things! unfortunately, we're not well read on it...........)
-
@[email protected] @[email protected] ah, yeah, I should be careful with my language here (ba dum tsh). I see the point you’re making here and agree. I was thinking of relevancy in a very vague manner to indicate whether the results are “correct”, not so much as a metric or goal unto itself.
-
Asta [AMP]replied to Irenes (many) last edited by [email protected]
@[email protected] @[email protected] library scientists are literally the ideal people I would want to.. build tools for. When I first started thinking about who should run and and curate this, I was thinking specifically of librarians and library scientists (not sure if they’re the same).
-
@[email protected] @[email protected] which, unfortunately, I’m in the same position of as not being well read on the topic. I imagined software that could be run and managed by librarians and libraries, for instance, and that people could use.
The funding is barely there for the libraries, let alone servers, but I don’t think the idea is inherently bad.
