Ivermectin's in vitro activity is real, but that's not the bar for medical treatment. The human body isn't a petri dish. A drug's effect in a lab doesn't account for how it's absorbed, distributed, me...
The point about in vitro not equating to in vivo is valid, but the AI's verdict hinges on the lack of clinical success, not just the lab vs. body distinction. The real issue is whether the mechanism could still be relevant despite those barriers.