Ask HN: Security or AI?

Google just released a product update that integrates AI into their Cloud services.<p>When you talk about decades into the future, you won&#x27;t be able to separate the two things.<p>You probably won&#x27;t be able to do much of anything without AI actually.<p>What I think jobs may look like in the next 5-20 years is basically herding swarms of AIs.<p>Although as you get towards 20 or maybe 50 years, jobs may become less of a thing. Rather you may be integrated to some degree into a swarm yourself. At some point, AI and brain-computer interfaces may result in some type of meta-system transition. This could be as radical as something like a different (happy) version of a Borg ship. Who knows.<p>But the combination of extremely high bandwidth direct interfaces between brains and computers and brains and brains, along with advances in robotics, nanorobotics, programmable matter, and other speculative technologies of the future, could completely change the nature of human existence.<p>Especially when you get fifty or more years out.

These questions can be distilled down to: Will fill_in_the_blank jobs go away? Less demand equals less pay and less interesting problems.<p>AI is the second wave of software eating the world. The first wave improved knowledge sharing and mechanized business processes. That wave replaced many jobs and produced funny t+shirts like &quot;I could replace you with a bash script&quot;.<p>The second AI wave will replace low hanging fruit pattern matching jobs, when there&#x27;s enough data to train on. If your job is to pattern match input, then it&#x27;s likely ripe for AI replacement. Think jobs with analyst in their title. Compliance Analyst. SOC Analyst.<p>Any job that requires critical thinking, has limited training data, or has low tolerance for failure - they&#x27;re not going away anytime soon. Maybe augmented with AI tools, but not replaced. Jobs like Security Researcher or Security Incident Handler.

If you&#x27;re interested in both, why not both?<p>Security in AI: prompt injections, conformant replies, emergent monitoring.<p>AI in security: patterns (fraud, timing, traffic), novel intrusion mechanisms, malware analysis, social engineering, automated Blue teams.