This reminds me of the Steve Jobs Dropbox story. "That's not a product, that's a feature."
Dropbox is still a company, sure. But it really is just a feature now. If you weren't locked into using it before its functionality became a feature, chances are you're not using it. No new company starts up and starts using Dropbox.
Yeah, these things can be products in the enterprise space where there you have different expectations for where data is stored, and where the problem space you are trying to fix are narrow enough that you can throw real money at it, but that doesn't translate to the consumer space.
This is also why apart from generalized enterprise needs like MDM, Apple doesn't engage with enterprise because the money is in being willing to address parochial needs, and Apple doesn't do that.
I'm sure ChatGPT will earn real dollars off the back of real estate agents who tire of writing MLS listings, but I don't see a consumer market for them. And the problem is that ChatGPT turns almost every other market into spam, which
inevitably gets a response to that spam. Maybe you have this mutually assured destruction of ChatGPT spamming resumes to employers and employers responding with AI countermeasures to read and sort those resumes, but like with every other market that falls to spam, the value of everything in that market falls to zero which makes it hard to then generate revenue off of it. At some point Zillow sees an opportunity to bypass the the MLS spam and monetize a non-spam alternative, cutting OpenAI out entirely.