Sure, but they were doing that stuff. They had ML people, infrastructure, marketing, branding,etc... already. Their product sucked, but they could have copy-catted OpenAI in 2022+ like everyone else.

I don't think that would have gotten them much of anywhere. They already spent a decade trying to find markets for Watson to fit and generally failing at it. The problem with Watson wasn't technology, it was that it had no direction.