Excuse my skepticism on that, but I observe just the opposite:
- Long buried issues surface and get resolved quickly;
- Test driven development is wildly augmented by the AI agents keeping the tests current;
- Experiments with solutions that you would never cause to yourself become daily throwaway effort,
to name a few on the reliability part.
Now security is another matter. We are, as of today, unprepared on the volume of issues that surfaces. Hence the model guardrails. But consider having dormant security holes, that were simply too complex for a human to uncover. That is just a finer version of security through obscurity.
Industry is changing rapidly, I would expect few incidents, few course corrections here and there, a sharper learning curve, but no cataclysm what so ever.
Comments
Excuse my skepticism on that, but I observe just the opposite:
- Long buried issues surface and get resolved quickly; - Test driven development is wildly augmented by the AI agents keeping the tests current; - Experiments with solutions that you would never cause to yourself become daily throwaway effort,
to name a few on the reliability part.
Now security is another matter. We are, as of today, unprepared on the volume of issues that surfaces. Hence the model guardrails. But consider having dormant security holes, that were simply too complex for a human to uncover. That is just a finer version of security through obscurity.
Industry is changing rapidly, I would expect few incidents, few course corrections here and there, a sharper learning curve, but no cataclysm what so ever.