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

ChatGPT and the Enshittening of Knowledge

castlebridge.ie

Post reply on HN
Print view View original

241–250 of 301 posts

Re: ChatGPT and the Enshittening of Knowledge

#241

Post by fnordpiglet »

banannaise wrote: post #212
fnordpiglet wrote: post #125

I think an easy way to identify transformative technology is how strong people’s reaction is against it. I remember similar freak outs about the graphical World Wide Web, smart phone, etc. Somehow things are not quite so bad and considerably better than predicted, while still having negative side effects. Maybe an outcome will be knowledge will be better structured, ala Wolfram Alpha, and just depending on random tex…

> I think an easy way to identify transformative technology is how strong people’s reaction is against it. There is also strong negative reaction to technology that does incredible damage to society. This is just an excuse to ignore people pointing out issues.

No, I don’t think so. I think I discussed that we would tackle the issues and some would be satisfied, some unresolved, and some ones we missed will emerge. I’m not saying the hyperbole cycle isn’t necessary or useful, I’m saying it indicates something significant. We couldn’t possibly resolve the issues identified if they weren’t identified to begin with. But I also think the bold predictions of what will go wrong are usually bold predictions of what could go wrong, and the very act of predicting it makes it less likely to actually occur. Finally, on a personal level getting bent out of shape about these things is pointless. We can identify challenges with new tech without it becoming a frantic exercise of hang wringing. Intentionally addressing issues knowing it’s happening regardless of your anxieties is a much more useful expenditure of energy, even if it doesn’t boost your blog/social media/impressions quite as much as hyperbolic hyperventilation.

Re: ChatGPT and the Enshittening of Knowledge

#242

Post by 93po »

fnordpiglet wrote: post #125

I think an easy way to identify transformative technology is how strong people’s reaction is against it. I remember similar freak outs about the graphical World Wide Web, smart phone, etc. Somehow things are not quite so bad and considerably better than predicted, while still having negative side effects. Maybe an outcome will be knowledge will be better structured, ala Wolfram Alpha, and just depending on random tex…

I feel like semantic web and OP are two independent topics. Semantic web is about how content is delivered. OP is about the content itself.

Re: ChatGPT and the Enshittening of Knowledge

#243

Post by venatiodecorus »

lubujackson wrote: post #17

If you think of the knowledge base of the internet as a living thing, ChatGPT is a like a virus that now threatens its life. This is the same process SEO spam caused for search - it hampers the nature by which things function and the river needs to reroute (pagerank then usage metadata) to replace the lost signal. ChatGPT is more of an existential threat because it will propagate to infect other knowledge bases. Luke…

> All signs point to this strengthening the value of curation and authenticated sources.

that sounds like a positive and greatly needed outcome

Re: ChatGPT and the Enshittening of Knowledge

#244

Post by forgetfulness »

ren_engineer wrote: post #44
lubujackson wrote: post #17

If you think of the knowledge base of the internet as a living thing, ChatGPT is a like a virus that now threatens its life. This is the same process SEO spam caused for search - it hampers the nature by which things function and the river needs to reroute (pagerank then usage metadata) to replace the lost signal. ChatGPT is more of an existential threat because it will propagate to infect other knowledge bases. Luke…

>And worse, then ChatGPT will digest its own excrement, worsening its own results further I wonder if we'll get a "dead sea effect" with AI, I've seen some stuff saying they've basically run out of high quality training data and now the training pool will get poisoned by AI generated shit. Basically garbage in, garbage out and these large language models might not be able to improve

Pre-ChatGPT datasets will become prized commodities, this sort of AI will be trapped in a stasis of pre-2023 pop culture as subsequent AIs will need to use datasets that hadn't yet been contaminated by pervasive ChatGPT spam.

It will be like low-background steel, steel that has somehow been isolated from atomic fallout form the mid XX century onwards, and must be used for radiation-sensitive equipment: https://en.wikipedia.org/wiki/Low-background_steel

Except somehow worse, because it's just steel, this is culture.

Re: ChatGPT and the Enshittening of Knowledge

#245

Post by miguelazo »

gptgpp wrote: post #219
miguelazo wrote: post #83

Earlier quoted context omitted.

It's telling that you only cited examples of scientific subjects. As the other commenter noted, articles of consequence for public debate (politics) are generally terrible and there are lots of "editors" who are working for deep state cut-outs doing nothing but trying to damage the reputation of intellectuals who are a danger to the status quo.

Yes I agree, it is telling. It is similarly telling that conservative wikis have barely any articles on core topics like engineering, mathematics, philosophy, and the sciences. These intellectuals you're describing oddly don't seem to have much interest in things most people would deem intellectual... For example, compare the Wikipedia article on Leonhard Euler with that of conservapedia... It's so absurd I had to do…

Wow, I didn't even know Conservapedia was a thing. Although conservative interests have way more money to throw around, so not surprising someone would sponsor such a dead-end project. Similar to their wiki directory of lefty intellectuals; can't remember the name.

Re: ChatGPT and the Enshittening of Knowledge

#246

Post by 93po »

brap wrote: post #221
manytree5 wrote: post #133

The internet has been groaning under the load of bots, automation and distrust for a while now anyway: in my mind what is needed is a system for proving constituent parts of your identity which is also privacy preserving. I'm thinking of the work around DID re: rebooting the web of trust, and the W3C working group If you can wield an ephemeral and verifiable token which asserts your humanness, hierarchically derived…

What’s stopping you from using this token with a bot?

I disagree with top comment a bit - I think we need tokens that are authenticated to an identity much like SSLs certs. If you choose to use your token to post AI generated crap on the internet, I can choose to block it. I can also choose to subscribe to community managed block lists. I think content on the internet needs to be reputation based, and if you choose to read content that isn't signed by a real person's identity, then it's sort of on you if you wind up reading bot written crap.

I think community managed lists would also be a great way to filter out people with annoying or problematic communication styles or opinions. If you publicly support Jordan Peterson, I can add you to the list of Jordan Peterson supporters, and as a community we'd never have to read anything you post anywhere, ever.

Re: ChatGPT and the Enshittening of Knowledge

#247

Post by miguelazo »

ROTMetro wrote: post #192
miguelazo wrote: post #77

Earlier quoted context omitted.

There's also tons of censorship on Wikipedia based on nothing but ideology. Just look at how the Grayzone can no longer be used as a source, based on claims it is "state-affiliated" media, despite ZERO evidence after literally years of such BS claims and now documented evidence of Western states targeting them (look up the exposé on Paul Mason) because they report inconvenient facts about what the security state is d…

Yes, the problem with Wikipedia is that it doesn't allow enough fringe, pro-violence conspiracy theorists! Nailed it!

That's an amazing take, since Grayzone is very explicitly anti-war, and one of the very few US publications taking an active stand against US proxy wars. If anyone is a pro-violence conspiracy theorist, it is the "paper of record" (NYT), which has actively supported/facilitated virtually every gruesome military intervention of the US for well over a century.

Re: ChatGPT and the Enshittening of Knowledge

#248

Post by miguelazo »

milsorgen wrote: post #196
miguelazo wrote: post #77

Earlier quoted context omitted.

There's also tons of censorship on Wikipedia based on nothing but ideology. Just look at how the Grayzone can no longer be used as a source, based on claims it is "state-affiliated" media, despite ZERO evidence after literally years of such BS claims and now documented evidence of Western states targeting them (look up the exposé on Paul Mason) because they report inconvenient facts about what the security state is d…

What state is Grayzone supposedly affiliated with? And what of Bellingcat, I have heard it's state affiliated. Is it an accepted source? What determines state affiliation? This kind of stuff has soured me tremendously on Wikipedia. Post it all and let me sort out I say.

They are all Kremlin (or previously Assad) assets, according to groups like Bellingcat, for which there is plenty of hard evidence of state control/funding.

Re: ChatGPT and the Enshittening of Knowledge

#249

Post by randcraw »

discreteevent wrote: post #15

> When we turn the job of creating the first draft of the thing over to ChatGPT we risk removing the “figuring shit out” part of everyone’s career-path. And we get away with that while we still have people who have figured shit out. That's the main problem from my point of view.

Agreed. The LAST use of ChatGPT should be the first draft. It's clueless about substance or the purpose of the essay, neither what's necessary nor essential. As such, its best use is to flesh out an outline that HAS these things into a essay or narrative that integrates them into a narrative that flows.

The trick to that is, how do you take away ChatGPT's inclination to build upon the wrong premises or the wrong final message, and instead confine it to the role of verbally adroit assistant -- so it only expands a skeletal design into an well written final product? Now THAT would be a use for ChatGPT I could get behind.

Alas, that wasn't the goal of its designers. By giving it almost total control over the plot line and requiring no fact checks or bibliography, the spewing of unchecked drivel was its only possible mission in life.

Congrats, OpenAI. You've automated spam.

Re: ChatGPT and the Enshittening of Knowledge

#250

Post by forgetfulness »

x86x87 wrote: post #183
pawelduda wrote: post #80

Internet often feels it's getting rotten with SEO spam with very little substance already. Now the flood gates are open to generate more shit in a much better wrapping. And it's going to take more effort to check its quality than to produce it. Unless we get better tools to do that - but text has more dimensions to process/understand as opposed to for example AI generated images which (for now) can be told apart by g…

What was the saying? It takes 10x more effort to refute BS than it takes to produce it? So if we make generating BS easier and that BS is fed for generating more BS we're going to get to a point where everything is BS. I call this the BS singularity!

We have finally automated the production of half-truth to a scale never before envisioned, maybe the limit will be the capacity for humans to absorb bullshit, it's a testament to how revolutionary this technology is that we'll be able to quantify just how much bullshit people can process over time, and if supply will finally outstrip demand for it.
Post reply on HN