Live data from Hacker News

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

thewalrus.ca

621–630 of 1001 posts

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

#621
post #412

Earlier quoted context omitted.

They are repackaging bunch of search engine results, including Yandex (and paying for it). Yandex is Russian. I would not describe it as Kagi being indirectly Russian.

> I would not describe it as Kagi being indirectly Russian. What do you mean? That Kagi is directly Russian? Paying a russian provider isn't against the law and at some point we need to bring Russia back into the warmth of the West.

Perhaps “at some point” is not “while they’re still waging a war of aggression”, though.

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

#622

I called this maybe 3y ago, but I think so did everyone else that was sane. Sure, we get immense value from AI, but indiscriminately injecting into everything, the one thing we know to be unreliable above the threshold we used to fire people for, is probably the greatest undoing of all the good companies like Google brought to the internet. I mean what a way to destroy your legacy of democratizing information. The am…

I find it so interesting how many comments about AI start with a disclaimer that of course it's very useful, before pointing out a negative of any sort. I do it too so it's not a criticism, but I find it kinda interesting culturally. It's a reflection I think of the environment that we've created for ourselves, or has been created for us (or some combo) where somehow you can't point out an unqualified negative of thi…

Well I think theres a lot of social pressure to disclaim "yes of course its useful" because if all you do is say "hey AI has this problem," you'll have AI bros replying to you telling you about all the good and saying you're misrepresenting AI.

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

#623

Earlier quoted context omitted.

AI is like eating cardboard.

AI can make fully functioning computing solutions. That's just a fact and denying it is like denying that a bicycle can roll and only four wheeled vehicles can. Or saying that a car is a better vehicle than a bicycle. That is probably true, but many times for many people a bicycle is all they can get.

Glad you went for this analogy. It’s like an AI producing a board with a beach umbrella nailed on it and four wheels, and saying “meh it’s good enough for transportation, human car designers are doomed”. It’s functional, right?

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

#624
post #569

Earlier quoted context omitted.

That was a temporary blip. Specific technologies facilitated an age of mass, top-down communications in the mid 20th century, just by virtue of how those technologies worked. It enabled a small group of people in charge of the radio/TV stations to have a monopoly on communications and media. The technology both before (printing presses with myriad newspapers) and after (the Internet) were by their nature decentralize…

Not totally convinced about the printing press example (expensive, technical). The biggest presses, with the longest reach (daily newspapers) were often owned by the richest people. What's unique now is decentralized + extreme potential reach. Or it was. Now we have to consider bias in the algorithm that sticks stuff in front of us and is, I suspect, the modern equivalent of those daily newspaper presses.

> The biggest presses, with the longest reach (daily newspapers) were often owned by the richest people.

You're assuming that "rich" is the only relevant dividing line, which it isn't.

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

#625

Earlier quoted context omitted.

LLM training will eventually transition from real data to synthetic data, same as alphago -> alphazero. AI companies are also working to integrate training with real-world experience through sensors and robotics, to shrink the gap between human experience and hallucinated LLM experience. They all have archives of pre-LLM content. There's also archive.org, google books, and pirate ebook archives. I don't know what the…

That (theoretically) solves training, but it doesn’t change the fact that even smart models can’t extract useful information from a dead internet, so you’ll always be stuck with a stale training cutoff. This is already a problem I run into a lot. I search something first. Top results are slop sites, so I switch to a chatbot. Its answers look suspiciously similar to the slop sites I just noped out of. Check the source…

And the training of future models will have to contend not only with slop, but also huge amounts of content specifically designed to "taint" future training data. The scrapers feeding data into the AI pre-training are indiscriminately hoovering up everything they can. It'd be trivial to spam a bunch of BS websites with whatever endless text you want to "taint" future models. Post tons of examples of insecure code, publish package.json files pointing to some malicious library, etc...

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

#626

I called this maybe 3y ago, but I think so did everyone else that was sane. Sure, we get immense value from AI, but indiscriminately injecting into everything, the one thing we know to be unreliable above the threshold we used to fire people for, is probably the greatest undoing of all the good companies like Google brought to the internet. I mean what a way to destroy your legacy of democratizing information. The am…

I called it 2001/2002 or whenever they appeared when I tried to explain why personalized search results are the beginning of the end of a shared reality and therefore the ability to reason and act in public, and with others. I bet some still consider it hyperbole. It's just taking in trends and seeing where the glacier moves to, how the cookie will crumble so to speak.

Personalization came along in MMOs around the same time, and it was hard to communicate the difference to people between a persistent MMO zone as a "place" where people and things exist, even if those things are inappropriate for your character or hostile to the point that you can't survive there or too crowded for the intended flow... versus an instanced MMO zone, a little pocket universe for you specifically that has no strangers in it, which is more of an "experience" or "cut-scene" for you to imbibe prepared content.

A heavily instanced MMO may as well be a single-player or co-op game with a shared chatroom as a lobby, there is absolutely none of the emergent play, none of the characterization of players & groups of players, nothing original. Instead, give me the same ground to walk on as everybody else, even if our tread wears all the grass off of it; Half of my life in these games was getting bored waiting for a mob to spawn and fucking around making friends waiting in line, or ganging up on the guy who wanted to cut in line, or waging war on the clan that currently controls the needed territory, or grouping together to fight mobs higher than I should be able to hit (which the instance will just Adjust Downwards). Instancing and other auto-personalization algorithms cut the annoyances that are central to human socialization, achievement loops, and wayfinding.

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

#627

Earlier quoted context omitted.

That was a temporary blip. Specific technologies facilitated an age of mass, top-down communications in the mid 20th century, just by virtue of how those technologies worked. It enabled a small group of people in charge of the radio/TV stations to have a monopoly on communications and media. The technology both before (printing presses with myriad newspapers) and after (the Internet) were by their nature decentralize…

It takes some technologies more time than others to succumb to the corrupting pressures of money, but they all fall eventually. We must keep outrunning it, building ever more resistant methods of communication. I'm hopeful that the next one will last even longer since we'll have to build it explicitly for resisting corruption. Neither the printing press nor radio nor webs 1 or 2 had fighting back as part of their DNA…

The top-down centralization of mid-20th century technologies wasn't caused by "money," it was caused by their technological nature. They inherently relied on limited radio spectrum that was doled out to specific companies by the government.

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

#628
post #622

Earlier quoted context omitted.

I find it so interesting how many comments about AI start with a disclaimer that of course it's very useful, before pointing out a negative of any sort. I do it too so it's not a criticism, but I find it kinda interesting culturally. It's a reflection I think of the environment that we've created for ourselves, or has been created for us (or some combo) where somehow you can't point out an unqualified negative of thi…

Well I think theres a lot of social pressure to disclaim "yes of course its useful" because if all you do is say "hey AI has this problem," you'll have AI bros replying to you telling you about all the good and saying you're misrepresenting AI.

I agree and I think there is also a more insidious meta aspect: for every comment like this there is the worry that some X multiple more people hold that view but won't say it.

Everyone else also thinks like that and so isn't normally unambiguously critical, so you end up not knowing who really thinks that way and who's just doing it because of the game theory.

It could even be practically nobody. There's no way to tell until the deadlock breaks.

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

#629

I called this maybe 3y ago, but I think so did everyone else that was sane. Sure, we get immense value from AI, but indiscriminately injecting into everything, the one thing we know to be unreliable above the threshold we used to fire people for, is probably the greatest undoing of all the good companies like Google brought to the internet. I mean what a way to destroy your legacy of democratizing information. The am…

I find it so interesting how many comments about AI start with a disclaimer that of course it's very useful, before pointing out a negative of any sort. I do it too so it's not a criticism, but I find it kinda interesting culturally. It's a reflection I think of the environment that we've created for ourselves, or has been created for us (or some combo) where somehow you can't point out an unqualified negative of thi…

To be fair that is also true to the Internet itself, it's very useful but has negative side-effects.

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

#630

Earlier quoted context omitted.

I called it 2001/2002 or whenever they appeared when I tried to explain why personalized search results are the beginning of the end of a shared reality and therefore the ability to reason and act in public, and with others. I bet some still consider it hyperbole. It's just taking in trends and seeing where the glacier moves to, how the cookie will crumble so to speak.

I have a dream, to ban targeted advertising globally. It would solve many problems and possibly, cherry on top, kill some corporations

Stop the purchase and sale of user data, period. Kill the weed at the root.
Post reply on HN