Everyone has the same tools now
One of the things on my mind now is: with AI and everything comes all these security questions and all these potential risks.
But the philosophy that I have around this is that even the legacy guys that feel like they're protected: if somebody who is well or ill meaning, and has access to even just Opus or Sonnet, points an agent at a particular system, you would eventually find a way in there. Even if it was a human that wrote the code 100%.
People have this mentality of, "Oh, AI ships bad code," as if human beings have been shipping perfect code before AI came.
The world that we're in now is that you almost have to think that everybody out there has all the most powerful tools. So how do you protect yourself from the fact that somebody's going to be using AI to do a few things?
Said out loud in a team meeting, 14 August 2026. Trimmed for filler; not rewritten.