Everyone has the moment when they realized there might be something wrong with Scott Adams. For me it was in 1997. I know this because I filed the strips away as a reminder. This is one of them, tracked down on the official web site.
Scott Adams allegedly worked as a programmer, so surely he must have heard of the Halting Problem? It's the idea that you might want to write some software that can analyze other software and determine if it will successfully run to completion. Alan Turing proved it's impossible to do that. So, it is absolutely possible to prove that something can't be done with software.
The Halting Problem is literally CS101 stuff. Did he ignore it so he could go on an anti-science rant, or was he in Dunning-Kruger territory even back then?