I think UNIX will be secure; next generation of AI will be able to scan directly the image of a compiled kernel and will be able to detect -- if not all -- harmful things.
Self-learning via books. I've bought a bunch. Am currently going through Susskind's Theoretical Minimum Classical Mechanics [1], also looking through No Nonsense Classical Mechanics [2].
AI chatbots sometimes come to rescue when I am stuck. Not with mathematics though.