Its going to be interesting.

They don't really have to follow the path of Google/Meta/Twitter etc which hardly had any customers/companies subscribing or paying for their services in the early years. That kind of sowed the seeds of turning them into Adtech companies.

Plus most people forget the collection of personal data started happening cause Google/FB/Youtube search was quite crappy without Context. As info exploded finding stuff got harder. Personal data collection was the answer. That accumulating personal data pile played a big role in making these companies very attractive to the Ad industry and the large corps behind them.

Open AI feels slightly different in that it has real revenue flowing in. For a wide range of subjects, it doesn't need as much personal info about the customer, to provide high quality answers. So more like Netflix in the early years. Paying customers. Good product.

Only diff is Netflix targets people's entertainment budgets, which is a small bucket of piss compared to Global GDP. Total addressable market for OpenAI meanwhile is looking really massive. It feels like it will be a long while before they hit some subscriber revenue upper limit and have to rely on Ads.

They are engaging with a lot of industries that could see their work generate immense value directly without needing to change the base services. If they end up being essential in cutting out half the costs in Clinical trials for drugs, they could easily pocket 10% while everyone around the table still being happy. And just imagine how much red tape they could potentially cut out the legal and political system.

And of-cause, their “moat” is pretty much just model training costs, but that slight up front capital cost is still peanuts for Microsoft, Google and Apple, so they can’t really afford to significantly degrade their services.