The skills I think we might gain (AI)

The general consensus, and common sense seem to dictate that the more we lean on LLMs for coding, the weaker our ability to code by hand will become. I feel like it's something all experienced engineers will be able to get back if they decide they want to put the effort in, but the AI genie doesn't look like it's going back in the bottle any time soon, and so the motivation to put that effort in is always going to be competing with the motivation for doing something else important.

That said, like many others, my output has gone up significantly - I would suggest 6x based on the number of projects I'm shipping compared to pre AI periods (and I don't think the quality has gone down like many suggest is inevitable), but I am tired. Agentic software engineering seems to come with fatigue, as numerous other blog posts I have read recount.

So my brain must be doing something ... right?

Decision Making, At Pace

It's like I've become and architect and engineering manager, with super keen, super fast employees. Those employees do need significant guidance, and help with 'taste' as it has been termed (I don't care for all of the talk about 'loops' yet. I've been wrong before, and I'll never say never, but that just doesn't look like a good way to build quality software to me).

And I am making decisions, and weighing options, at pace, all the time. I'm using Matt Pocock's Grill With Docs Skill extensively, and it is thorough, to say the least. The result is solid well designed software that does what you designed and everything you hadn't thought of, but the sheer number of choices that you need to make is exhausting.

But neuroplasticity is thing, and I expect that while my code grinding skills weaken (hopefully 18 years experience means something will stick around for a while), my decision making, intuition, forward planing, problem description and design skills will improve.

This is something that's going to be hard to measure. For now I'm watching for two things. Software quality, and the success of projects against their stated aims.