"Attention is all you need"

True! AI will make skills and focus more valuable.

Nothing beats intuition + experience

I am waiting for Cal Newport's Deep Work to become a managerial trend like Agile did. And of course it will be weirdly twisted in practice. We'll see.

> This reliance on readily available solutions, particularly for familiar problems, creates a real risk: engineers may inadvertently atrophy their own problem-solving skills, hindering their ability to tackle truly novel challenges.

Yes, that will happen. But it also happens every time we move up the abstraction ladder. Most engineers go through their entire careers and never do anything TRULY novel.

For me, there's a sharp binary: If I ask AI to "own" a coding-problem solution — with me passing back the failure responses until resolved — my mind gets numb and I learn nothing. If I insist on owning the solution — using AI in my effort to better understand the problem space — my mind is active and I get better at coding. Sometimes I'm lazy and fall into the former. But mostly, so far, the latter.

> But, like any tool, LLMs should be used wisely.

While it's difficult to define, wisely can turn 'LLMs are useless' to 'ten X productivity boost'. However, at the end, of course, it all comes down to products. Before LLMs stole the show, we had built beautiful system software over the course of decades, linux, git, k8s and rust and yet the products that we use everyday are mostly (mostly) user-hostile and incorporate dark patterns, offer a suboptimal UX, and (in my opinion) sometimes involve outright inhuman marketing practices. That being said, even if you get AGI I don't think it will lead to any breakthroughs if we continue to do 'software engineering' like this year after year.

The skill of the future is behaving ethically.