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

As AI eats the web, the internet’s collective memory is disappearing

thewalrus.ca

Post reply on HN
Print view View original

301–310 of 1001 posts

Re: As AI eats the web, the internet’s collective memory is disappearing

#301

Post by kees99 »

neuralkoi wrote: post #87
paxys wrote: post #40

Earlier quoted context omitted.

All the information Gemini surfaced was created with human effort and published on the internet with the expectation that humans would visit the website and the creator would get some reward - advertising dollars, bragging rights, popularity, subscribers or whatever else. If the only visitors to websites are now LLM training bots then what incentive is there to publish anything new? For how long can we continue to re…

I've seen websites put up some draconian measures to try and get a grip on the scraping. So much for the sub-second loading experience when you have Cloudflare, Google, Anubis, and all these other captcha services trying to see if you're a human. It's made the web browsing experience so much worse. Some of the proposals to address this include charging bots for access to web resources, but they will also have repercu…

> Cloudflare, Google, Anubis, and all these other captcha services trying to see if you're a human.

Yep. IMO, this is so far the biggest AI-inflicted damage to the web. A bit of anecdata - wikipedia (and all other wikimedia sites) are blocking my Firefox since about a week, with a "please respect our bot policy" message. Outright block, not even a captcha.

It took me a while to figure out they don't like me disabling some SSL ciphers, so now "JA4 browser fingerprint" is not matching user-agent. Funnily enough curl (what I would imagine a bot would use) pulls exact same URLs from exact same client IP, just fine.

Re: As AI eats the web, the internet’s collective memory is disappearing

#302

Post by root-parent »

ChiMan wrote: post #196

AI will kill the internet because it is killing the incentive to make it. It is an industrial-strength example of why we don’t allow stealing.

And that will kill AI itself, since much of what it knows is from what learned from StackExchange before this latest one demise.

And before the obvious comments on how GenAI is creative, then please do this OpenAI and Anthropic, for your next LLM. Just teach it Python, C and Rust and give it some good books. But dont give it access to Github...lets see what you can do then...

Re: As AI eats the web, the internet’s collective memory is disappearing

#303

Post by sph »

sixtyj wrote: post #287
hliyan wrote: post #213

Earlier quoted context omitted.

Recently, I had some ideas I would normally just put up on a blog, in the public domain for anyone to develop on top of. Now, I'm feeling slightly reluctant because an LLM will ingest it, remix it and serve it in response to a query by some unimaginative individual who will either conclude that they are smart, or that LLMs are capable of original thought, or both. And they will have no clue where the idea originated…

Just do it for yourself. I stopped worrying about who’s gonna use it. 20 years ago I thought that we will wait to have kids as “a war will come”. I was overthinking and I am happy now that I changed my mind :) Is it useful or does it make you happy? Or it might even make some money? Nice. Do it. My life is easier.

That's the same conclusion I reached.

I'll keep publishing static websites, so I don't even have to worry about load and CPU usage. I don't care who reads it, the value for me is in writing.

I still haven't changed my mind on the "shall I have kids" problem :P

Re: As AI eats the web, the internet’s collective memory is disappearing

#304

Post by zelphirkalt »

sph wrote: post #277
ChiMan wrote: post #196

AI will kill the internet because it is killing the incentive to make it. It is an industrial-strength example of why we don’t allow stealing.

More than a facilitator of theft, LLMs are the tragedy of the commons at industrial scale. The public internet is dead, the future is private invite-only walled gardens. Corporations love a walled garden, what we need is open-source frameworks to create these islands, rather than defaulting to horrible systems like Discord and Twitter-clones.

To understand your comment correctly: What does "these islands" refer to? Walled gardens that we create ourselves using the open source frameworks?

Re: As AI eats the web, the internet’s collective memory is disappearing

#306

Post by pjc50 »

account42 wrote: post #279
pjc50 wrote: post #245

Earlier quoted context omitted.

Eh, part of what made early-internet so good was exactly that it did allow copying by users ; the DRM era was later. But it's a very good example of why not to allow for profit copying, because that absolutely will crowd out the original. Piracy has to exist at the margin. The zero piracy world would also eat its memories because none would leak into archives. Remember Qubi? It wasn't even popular enough for people t…

I think the lack of credit is even more egregious and a bigger problem than the commercial copying.

Yes. Some communities are weirdly against giving credit or keeping the credit (e.g. cropping off signatures from artwork), which I've never understood. It costs nothing.

Re: As AI eats the web, the internet’s collective memory is disappearing

#307

Post by jfyi »

chrisjj wrote: post #240
t0bia_s wrote: post #223

Earlier quoted context omitted.

It depends on filed. As documentary photographer it motivates me even more to capture authentic images of life around me. I don't care about remixing, because that is not what makes documentary photography valuable.

And how is any future viewer going to tell your images from "AI" fakes?

With photos, I could see a cryptographic solution. Of course it would still need some kind of centralized trust, but it's doable if people cared enough. It could be applied by cameras themselves.

Re: As AI eats the web, the internet’s collective memory is disappearing

#308

Post by BrucecarlL »

root-parent wrote: post #302
ChiMan wrote: post #196

AI will kill the internet because it is killing the incentive to make it. It is an industrial-strength example of why we don’t allow stealing.

And that will kill AI itself, since much of what it knows is from what learned from StackExchange before this latest one demise. And before the obvious comments on how GenAI is creative , then please do this OpenAI and Anthropic, for your next LLM. Just teach it Python, C and Rust and give it some good books. But dont give it access to Github...lets see what you can do then...

[dead]

Re: As AI eats the web, the internet’s collective memory is disappearing

#309

Post by bko »

ChiMan wrote: post #196

AI will kill the internet because it is killing the incentive to make it. It is an industrial-strength example of why we don’t allow stealing.

I don't know how long you've been on the internet but the incentive to create new and original content was never that strong. Simple search terms return super-spammy websites (especially on mobile where ad-blocking is harder). SEO results for everything like simple search are awful, almost unusable. There hasn't been an incentive to create new original non-monetized content for the web for a while.

I trust LLMs more than search engines to discover my content and propagate it to users. They might "steal" something, sure, but I'm essentially invisible to the search engines as I could never hope to break into the top 10 links on a popular search term. LLMs can scan thousands of links and (for now) are more interested in quality rather than click monetization or referral incentives.

Re: As AI eats the web, the internet’s collective memory is disappearing

#310

Post by sph »

zelphirkalt wrote: post #304
sph wrote: post #277

Earlier quoted context omitted.

More than a facilitator of theft, LLMs are the tragedy of the commons at industrial scale. The public internet is dead, the future is private invite-only walled gardens. Corporations love a walled garden, what we need is open-source frameworks to create these islands, rather than defaulting to horrible systems like Discord and Twitter-clones.

To understand your comment correctly: What does "these islands" refer to? Walled gardens that we create ourselves using the open source frameworks?

It's still a work in progress of an idea.

I'm thinking more like mesh networks. I spoke of Reticulum elsewhere in this thread, but here I'm thinking I'd like the ability to easily join multiple TCP/IP networks (islands of connectivity) by social group (my friends) or by interest (pirate file-sharing group, my work intranet, a knitting community with their own IRC server, FTP, etc.).

Basically easy-to-use private & encrypted LAN overlays on top of the public internet. Each operator decides who to allow in or kick out of the network.

Wireguard solves the most of technical challenges, but it needs a frontend. The biggest concern probably is most software broadcasts their stuff across all interfaces, defeating the point of isolation between networks.

Post reply on HN