I think this article is making a pretty big assumption: that people making things with AI are also going to be publishing them. And that's just the opposite of what should be expected, for the general case.

Like I've been making things, and making changes to things, but I haven't published any of that because, well they're pretty specific to my needs. There are also things which I won't consider publishing for now, even if generally useful because, well the moat has moved from execution effort to ideas, and we all want to maintain some kind of moat to boost our market value (while there's still one). Everyone has reasonable access to the same capabilities now, so everyone can reasonably make what they need according to their exact specs easily, quickly and cheaply.

So while there are many things being made with AI, there is ever-decreasing reasons to publish most of it. We're in an era of highly personalized software, which just isn't worth generalizing and sharing as the effort is now greater than creating from scratch or modifying something already close enough.

> a company shared their work with the public for however long, then decided to leave what was shared up

More like a company took advantage of a community that expected their freely offered labor to not be commercialized at any point in time without making available said works in a fully free vector as well, as that's an implicit expectation behind "open source".