Are large language models a threat to digital public goods?

I think that is somewhat off topic. I don't see why rent seeking via IP laws should be bad while doing it via provision of AI wouldn't be.

> They monetised with ads and affiliate links

I consider this to be a problem on its own, but it's not relevant here because:

> In LLM land, they get no monetisation any more because nobody visits their sites, instead the LLM just regurgitates the answers they found.

That can't possibly be true, because if it were, there wouldn't be any travel blogs anymore today. All that travel spam has been made redundant approximately around the time Flickr was created, and every interesting location ever has been photographed from every interesting angle in a way neither me, nor you, nor your favorite travel blogger could ever hope to match. All the information they post has also been posted many times over by travel bloggers that came before.

The point being: travel information and photography is worthless commodity these days. Travel bloggers (or Instagrammers, or whatever) are not in the business of selling information. They're selling dreams and personal experiences. The photos and information are necessary as delivery vector ("social object"), but by themselves are worthless and not the point. The point is entertainment, and ideally getting you trapped in a parasocial relationship with the travel blogger/grammer, which gives them a recurring revenue stream.

> In LLM land, they get no monetisation any more because nobody visits their sites, instead the LLM just regurgitates the answers they found.

It's the same model as with most other ad-monetized social media publishing. People will keep visiting them for the same reason they visit them now, and for the same reason they have their favorite youtubers and tiktokers. LLMs and other generative models don't change anything here, at least not short-to-mid-term, because they can't convincingly replicate human connection and keep it up for long.

(Also, I personally don't buy that travel instagrammers can actually sustain their travel lifestyle through ads and affiliate marketing and sponsorship deals. I suspect most are funded in some way, whether by family wealth or by services performed while traveling around.)

> The search engines actively supported these authors, by sending them people who needed the answers they had. (...) A LOT of the useful information on the web was built on similar feedback loops and they go away in LLM land.

Hard disagree. The only feedback loop this created in practice is the one that displaces quality information from the Internet - the combination of SEO and ad-based monetization means the most scummy players are the ones with most money to stalk every conceivable search query. The results speak for themselves: making a Google query for pretty much any topic of interest to general population will give you only content marketing sites - results that carry negative knowledge, as in if you waste your time reading them, you'll come more misinformed about the topic than you were before. If LLMs make all that go away, I'm 100% for it.

As for "A LOT of the useful information" - nope, can't think of a single case where ad/affiliate-supported site was a good information source, vs. just displacing a better free source.

Why would they suddenly paywall information if they weren't already?

The way I see it, there are roughly three groups of information providers:

1. Those who do it pro bono - because they feel like its a worthwhile thing to do, or because they believe in by "pay it forward", or otherwise because they haven't even thought that what they share is worth trying to extract rent from.

2. Those who do it "for free", as a way to lure people to where they can expose them to ads, affiliate marketing, upsells, or other such schemes - making money by being predators using information as bait.

3. Those who just put up a paywall, being up front that they're selling information, not giving it away.

(There's also a weird "in-between" group of publishers that are almost like 1., except they're being funded out of marketing budgets of companies that figure providing quality information is good advertising.)

LLMs don't change anything for group #1. They may compete with group #3, but that's business as usual, not anything transformative. The group that's directly affected is #2, which also happens to be the group that produces all the garbage on the Internet, so I'm actually very happy to see them forced to find a more useful way of making money. Since group #2 produces "information" that's arguably negative knowledge on the net, it's likely to improve the amount of quality information you'll be able to find on-line.

LLMs aren't just a mere database containing indexed copies of other peoples' IP. AI companies are charging you for access to a sophisticated automated reasoning system, that necessarily had to memorize half of the Internet in the process of becoming capable of (some approximation of) reasoning.

(BTW. that you can even make a system this way is a huge breakthrough that's not being talked about enough.)

But even if they were a mere database indexing copies of other peoples' IP, then - copyright issues notwithstanding - the de-bullshittifying of information retrieval process alone would be service worth paying a lot of money for.

I mean, this is more-less what a student does when writing a paper, when they're forced to cite their sources. They first come up with an idea based on their own understanding/recollection, then they try to figure out where did they first took that idea from. If they remember a specific source, they'll cite that; if they don't (because there may not be one specific source they learned from), they'll search for some existing work that expresses the idea in question, and cite that.

I.e. in case of both the student and an LLM, correct citation doesn't actually mean the idea originates from the cited work - only that the work contains this idea.