Major Publishers Concerned Over LLM Training and Search Engine Traffic

[engine]

Major publishers (and all of us) have been concerned for some time over the rise of LLMs, training, and, ultimately, where the market is moving. Any publisher that creates something unique has been compensated in some way in the past with traffic, or Google's News Showcase programmes around the world.

The tipping point for everyone is the LLM AI answer system where a LLM is trained on content from the Net, and yet there's appears very little recompense for providing the training that generates the answers.

We all know this has become a challenge over quite a while.

Even Google knows this is a problem, and is in discussions with major publishers on this evolving aspect of the Net.

I'm not sure anyone has the perfect answer, but it's clear that there are pent up problems. Google's huge push to keep making money through all its moves to cover the SERPs with ads and sponsored listings. It's become ARPs (Advertising Results Pages).

Think about it for a moment and you'll see that the LLM AI output isn't, as yet, easily monetizeable, so Google, Bing and others have its own challenges.

For anyone providing input to train the LLMs and GPT models where is their payback? There's already been a shift that many publishers have blocked the training bot. Here's some of the topics here on that.
[webmasterworld.com...]
[webmasterworld.com...]
[webmasterworld.com...]

Those concerns relate to web traffic, whether publishers will be credited as the source of information that appears in the SGE summaries, and the accuracy of those summaries, those publishers say. Most significantly, publishers want to be compensated for the content on which Google and other AI companies train their AI tools – a major sticking point around AI.

A Google spokesperson said in a statement: “As we bring generative AI into Search, we’re continuing to prioritize approaches that send valuable traffic to a wide range of creators, including news publishers, to support a healthy, open web.”

On compensation, Google says it is working to develop a better understanding of the business model of generative AI applications and get input from publishers and others.

[reuters.com...]

[tangor]

Perhaps we'll see a return to old pulp rates of "pennies per word"...?

[engine]

I'm more concerned about traditional traffic drivers, or, as it is now, the lack of traffic to publishers, and our sites.

There has to be a way, and at the moment, the search offering is no longer offering traffic, or any recompense for training LLMs.

The whole point of an answer service is that it should be able to show the source of the answer.
Take a look at books: They reference the source of the research information

Why can't Google and Bing do exactly the same thing! They can, of course. In fact, they can do it way better than references in a book. [webmasterworld.com] It should be technically feasible to build links into the answers. Complexity isn't required. Keep it simple.

Monetizing it is still a challenge. The current model is way too heavy on ads, and distractions for a user, imho. I was looking at a search today and thought, oh no, where is the organic?
No wonder the traffic has declined.

For any business the investors will expect something in return, and that includes search engines, or advertisers.

I'm not sure the search services have the solution just yet while in this early development phase. Clearly, with the current model they will have to have ads and sponsorship in some shape or form, and we're watching the search services decline from what we've known.

I'd call for the return to a traditional search service, with options of answers.