I think your reply did a couple things:

* strengthened the point of the GoF comment

* pointed out how high-biocontainment is the key here, which obviously wasn't present in this case (and I'd say the vibes for AI research in general are that it's not high-containment)

* correctly argued that AI research is definitely worth doing and holy shit we need to work on alignment