Yes, but also viruses do not actually have goals as such - maybe 'replicate' could be their only goal. So a virus _could_ come up with a mutation that kills all

Yes, but also viruses do not actually have goals as such - maybe 'replicate' could be their only goal. So a virus _could_ come up with a mutation that kills all its hosts, it just dies out as well in the process. Not very reassuring for the host! As a slight aside - since I was talking about the weakness of biological analogy on HN recently - the way humans fly is very very different to how (most) birds do. If nanomachines actually could be built by humans (non-biological nanomachines) they might be as different to virus/cells/organisms as plane wings are to bird wings. Of course, also please do not prompt a frontier model to make nanomachines, either, thankyou.