The one possible hope here is that since these things started as paid services, we know subscriptions are a viable and profitable model. So there's a market force to provide the product users actually want, which does not include ads.
If OpenAI or the other players are pushed toward expanding to ads because their valuation is too high, smaller players, or open source solutions, can fill the gap, providing untainted LLMs.
Why wouldn't a company monetize both ways? Paid video streaming services still show ads, and when I pay for a movie in theaters, they're still doing product placements.
It makes the service worse. I won't pay for a streamer that uses ads. If they start doing that I'm out. Ofc, that doesn't mean it's not a net win for them across all customers, but it does mean there are a subset of customers who are now willing to pay for a different service, and the market has an incentive to service that.
Is there any reason to believe the current subscription models are viable and profitable outside of huge burn rates? Uber rides are now 4-5x what they cost when they were starting up, but uBer was disrupting an entrenched market with extremely high prices. Even today Uber's are still typically less than 50% of what a Taxi previously cost.
If LLMs are disrupting search then they would have to adopt a similar monetization strategy to be profitable. The major issue with that is LLMs are many orders of magnitude more expensive to run that a search engine.
If OpenAI or the other players are pushed toward expanding to ads because their valuation is too high, smaller players, or open source solutions, can fill the gap, providing untainted LLMs.