I agree partly with your comment, but I want to add this perspective:
The dot-com era treatened and killed many jobs in banking (bank tellers and such). AI is now doing the same, but now it is threatening the jobs of consultants.
So the author tried to fix a non-issue of somebody else's dog with human-grade probiotics.
I think that actually there was no issue at all and he just wanted to YOLO it.
And then he only goes to discuss if it was ethical as he didn't ask the dog? What about just not medicating a dog when there is no medical problem? What about consent from the owner? So much more issues with this than what the author discusses.
Why release a model with such strict filters? It would be more sensible from a buisness perspective if they had sensible filters in place right away.