@m-j-r.eth
while there should be a forcing function for improving all security at all times...
sensationalizing the 1970s concept of a computer worm by simulating AI against purposefully weak test environments does a grave disservice to that public interest. this kind of research is far more instrumental against democratizing AI and for the runaway ability of a capable minority to exploit more systems over time.
https://palisaderesearch.org/blog/self-replication