I'll bite because I find this compelling: if Apple is trying to build momentum for its Private Cloud Compute (https://security.apple.com/blog/private-cloud-compute/), it makes sense to try and get ahead of anything that could enable a competitor to build something similar. I'd think OpenAI getting billions shoveled their way in future investment just to throw darts at building any kind of moat would qualify, because for all Apple knows, OpenAI might eat their Private Cloud Compute lunch.

Granted I'm not sure there's much of anything to substantiate this, but I imagine Apple would know better from competitive intel.

(I feel a bit bad that my mind also immediately connects "Gruber" to "Apple" in less-than-savory ways, but alas, that's the reputation he built for himself.)