The tricky thing for Google will be to do this and not kill their cash cow ad business.
My guess is that Google/OpenAI are eyeing each other - whoever does this first.
Why would that work? It's a proven business model. Example: I use LLMs for product research (e.g. which washing machine to buy). Retailer pays if link to their website is included in the results. Don't want to pay? Then redirect the user to buy it on Walmart instead of Amazon.
People who are aware of that and care enough to change consumption habits are an inconsequential part of the market.
You combine this with Apple pushing on device inference and making it easy and anything like ads probably will kill hosted LLMs for most consumers.
This is a new line of business that provides them with more ad space to sell.
If the overview becomes a trusted source of information, then all they need to do is inject ads in the overviews. They already sort of dye that. Imagine it as a sort of text based product placement.
You might think that's the correct way to do it, but there is likely much more to it than it seems.
If it wasn't tricky at all you'd bet they would've done it already to maximize revenue.
It never will. By disincentivizing publishers they're stripping away most of the motivation for the legitimate source content to exist.
AI search results are a sort of self-cannibalism. Eventually AI search engines will only have what they cached before the web became walled gardens (old data), and public gardens that have been heavily vandalized with AI slop (bad data).
Disclaimer: google stock holder.
This is not for Google to decide.
The users have spoken clearly that (when given an option) they will not tolerate or succumb to the SPAM of shitty SEO-optimized content-farms which has been plaguing the internet for the last decade.
If Google don't provide meaningful results in their search page, people will use ChatGPT or something else to sidestep the SEO SPAM issue all together.