If this past weekend has proven anything, the state of OpenAI and the dependence newly developed products have placed on it is not as rock-solid as OpenAI and these products would like you to believe. That is not to say that AI, in general, is not worth exploring and building with, but it should not be the core of your product offering when you are not the company developing the AI technology.
Simply put, AI should be the seasoning that enhances your core product, not the meat of your core product if you don’t have visibility and control over it.
For years now, we have always known about platform risk. This is a perfect example of how you could do everything correctly; the platform has no issue with what you are building on top of theirs, and everything still falls apart not because of anything you did but because of decisions made by the platform company. In this instance, the OpenAI Board. Because you baked this third-party platform into the core of your product, you are now in a position that requires you to re-evaluate everything.
A more appropriate use of third-party platforms, like OpenAI, would be to use their technology to enhance yours only after you have functionality that can stand on its own. If OpenAI closes its doors tomorrow, could your technology continue functioning without it? I hope you can answer yes to this question.