Coding agents are here. So what?

Anyone with access to coding agents can spit out code. But not everyone using these tools can generate code that meets requirements without issues. It's not smooth sailing always. You don't want your application to have security issues, crashing in the middle of the night or some other weird behaviour.

So there's a gap here. Those who can effectively use these tools to solve the problem and those who can't.

To make efficient use of these tools, one must have the ability to evaluate the code generated by these coding agents, be in a position to ask (right) questions to steer the agent in directions it never thought about, uncover the edge cases etc. And all of this is possible only if you know what questions to ask and can think of in advance how the code is going to work, all the paths it's going to take etc. without even running it. This skill can only be demonstrated with a good understanding of how computers work in general and the underlying domain you are operating in.

It's not a new thing. It's been one of the most important skills before coding agents are here and it's going to be even more important skill going forward. So nothing changed?