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

71–80 of 1001 posts

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

#71

Post by penneyd »

paxys wrote: post #40
comrade1234 wrote: post #7

I spent the last three days (off and on) using Gemini to configure my edge router 4 with my iOS devices on a vpn and it's been awesome. In the past I'd do a google search and read a few sources of documentation, do another google search and read another set of documentation. Now, Gemini aggregates multiple pages together so all of the work of reading source docs from multiple locations is now n a single step. Oh, I s…

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…

Well in the example above the manufacturer still has incentive to provide the manual's and guides that describe how to use their products, and if that is subsequently served by an LLM that's totally fine. The only sites that LLM's would have a negative effect on are those that are only hosting content for the ad views.

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

#72

Post by indigodaddy »

comrade1234 wrote: post #7

I spent the last three days (off and on) using Gemini to configure my edge router 4 with my iOS devices on a vpn and it's been awesome. In the past I'd do a google search and read a few sources of documentation, do another google search and read another set of documentation. Now, Gemini aggregates multiple pages together so all of the work of reading source docs from multiple locations is now n a single step. Oh, I s…

Believe me they're making money off you one way or the other

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

#73

Post by t-writescode »

dylan604 wrote: post #50
t-writescode wrote: post #36

Earlier quoted context omitted.

That is an option for sure, but a very disappointing one. And it comes with a meaningful danger of ending up with 2 subscriptions: one yearly and one monthly. They really, really need to finish their “pay as you go” system. I don’t know why it’s not done yet and there’s been no word about it to my knowledge.

Maybe they should just enable a tip option? What do donations do to a company's tax liability and would it be worth their effort to enable something like that?

They already have a system for “prepaying”, but it can’t be used to put more credits into your account, they just sit there, waiting to be used by a future subscription.

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

#74

Post by Havoc »

paxys wrote: post #40
comrade1234 wrote: post #7

I spent the last three days (off and on) using Gemini to configure my edge router 4 with my iOS devices on a vpn and it's been awesome. In the past I'd do a google search and read a few sources of documentation, do another google search and read another set of documentation. Now, Gemini aggregates multiple pages together so all of the work of reading source docs from multiple locations is now n a single step. Oh, I s…

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…

That is next earnings quarters problem is the approach being taken

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

#75

Post by queenkjuul »

ElProlactin wrote: post #60
bitwize wrote: post #35

Earlier quoted context omitted.

On the other hand, when modern archaeologists discover "Claudius has a small dick" graffiti on the side of some God-forsaken wall in Pompeii, they're fascinated. The presence of such graffiti adds color and texture to the civilization inhabited by Virgil and Ovid. What's just disposable background noise to us may provide context into how we lived and thought to our far-future descendants.

> The presence of such graffiti adds color and texture to the civilization inhabited by Virgil and Ovid. It does, but do you think that people at that time thought anywhere near as much about preserving their scribbles as we do? I'd venture a guess that we've created more "content" since the advent of the internet than in all of human history prior, and most of it is stored on things that aren't even designed to last…

I think it isn't that we should save all of it, it's that we are not saving any of it. Random letters, notebooks, calendars, family photos, restaurant menus, etc have all proven useful to various historians, of which there will essentially be none from our era.

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

#76

Post by Eddy_Viscosity2 »

penneyd wrote: post #71
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…

Well in the example above the manufacturer still has incentive to provide the manual's and guides that describe how to use their products, and if that is subsequently served by an LLM that's totally fine. The only sites that LLM's would have a negative effect on are those that are only hosting content for the ad views.

Manuals don't always well explain how to use their products with everybody else's products because there are too many to do that. But there are lots of people trying things out and might figure out the fine details on how to make various things work. They then publish these how-to pieces (which exist no where else) to the internet, or at least they used to when there were incentives to do so.

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

#77

Post by swed420 »

Terr_ wrote: post #37
bloaf wrote: post #12

Earlier quoted context omitted.

As someone who regularly reads things online, then wants to read them again like 3 years later, Google has been monotonically declining in quality. Also I realized the other day how hard it is to find song lyrics for anything other than quite mainstream songs.

It used to be I could always find the quote I wanted from a book. Nowadays I need to keep my own copies...

> Nowadays I need to keep my own copies...

https://github.com/asciimoo/hister

> Hister is a private search engine for the pages you visit and the files you keep. It indexes their full contents so you can find information again from the web interface, terminal, or an AI assistant connected through MCP.

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

#78

Post by verdverm »

novafunc wrote: post #5

I occasionally use Google Search when DuckDuckGo fails to give me relevant. Almost always, Google has better results. Though I can find its AI answers annoying aggressive. I'll look up like two search terms and the AI will bullshit multiple paragraphs out of despite having zero context of what I am looking for. DuckDuckGo seems to have detection of whether it should give an AI answer. And it allows you to have more g…

that control (so I can turn it off completely) is why I picked DDG to replace Google, who force feeds us the hallucinations

I've stopped using DDG now because of result quality. I now use a "meta" search backed by EXA, Tavily, and SearXNG in parallel. It can be agentically de-dupped or summarized as needed. Search as we knew it is done, largely because clicking through to evaluate result relevance before diving deeper sucks. Now we have agents that can do that portion and perform multiple searches, building on information in the last batch, to collect good results

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

#79

Post by swingandamiss »

paxys wrote: post #40
comrade1234 wrote: post #7

I spent the last three days (off and on) using Gemini to configure my edge router 4 with my iOS devices on a vpn and it's been awesome. In the past I'd do a google search and read a few sources of documentation, do another google search and read another set of documentation. Now, Gemini aggregates multiple pages together so all of the work of reading source docs from multiple locations is now n a single step. Oh, I s…

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…

Gemini can just consume the device documents. There's an incentive for device makers to publish this content.

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

#80

Post by satvikpendem »

paxys wrote: post #40
comrade1234 wrote: post #7

I spent the last three days (off and on) using Gemini to configure my edge router 4 with my iOS devices on a vpn and it's been awesome. In the past I'd do a google search and read a few sources of documentation, do another google search and read another set of documentation. Now, Gemini aggregates multiple pages together so all of the work of reading source docs from multiple locations is now n a single step. Oh, I s…

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…

People write and create regardless of profit motive, it has been that way for thousands of years.
Post reply on HN