The AI Takeover: What physics, markets, and security professionals actually know compared to “the hype.”
This seems to be the thing most people are asking me about these days, either those who know I’m in Cybersecurity or those who find out, questions about AI soon follow.
Let me get my biases out of the way upfront. I am not afraid of AI taking my job, and I do not necessarily lie awake nights worrying about something like Skynet. But there is a very real part of me, that says to all of us (humans) that if AI takes over in such a way, on one hand, we all kind of deserve it! We do very little about such things because we love the conveniences they bring, and then we get upset when it’s something other than what we thought. Well, we should start learning some lessons about our complacency and addiction to convenience, but we don’t like hearing that.
However, whether you explicitly use AI or not, you implicitly have been for some time by obliging the “convenience market” and certain “features” on your smartphone (why do you think it’s called that?) and other smart devices, as well as free email services, etc. Features you can’t imagine living without at this point. Some of them extremely useful.
Right next to that,, I have also spent the better part of a decade watching machine learning quietly become the backbone of security operations: anomaly detection, automated triage, adversary emulation. In my field, at the enterprise level (which I am not in), AI is not a scary newcomer. It is a colleague. And it is largely welcome. But the current AI movement, the hype cycle, the market mania, the rush to embed autonomous agents into everything from enterprise software to your grandmother's laptop, is a story being told badly. It is being told by people with enormous financial incentives to exaggerate what AI can do, downplay what it costs and what can happen with it, and scare everyone else into accepting a regulatory framework that happens to benefit the tellers.
This post aims to pull together some simply but vastly sourced perspectives on what AI can actually do, what physically constrains it, what the market is doing to the narrative, and where the genuine danger lies for cybersecurity, particularly on the personal computing frontier.
