I think people are getting overly impressed/intimidated by tons of bugs being found by LLMs in a bunch of code that hasn't been looked at by more than a couple of people in years, or even at all since it was written. Those are going to run out. There won't be any code left that hasn't recently been looked over by an LLM.
Plus, the definition of the "easily found stuff" is a moving target. The AI models aren't static either. What takes a human reverse-engineer a week of deep insight today might just be a standard automated API call by 2027.
So while I would love for the dust to settle in a year, I think we are just looking at the new normal.
Thanks for reading the post and for the great counter-point!