I often think back to the early GPT days before agents as maybe the last time I will deeply learn a technical subject. I think that was when I actually learned the most as I had to stay completely in the loop. I had recently started a new job and had to work with k8s for the first time. Using GPT to help me implement new services and help me diagnose and fix issues with kubectl taught me so much.

Now with agents I don't have any insights into what its doing with kubectl and would have no reason to learn how it works (other than my own curiosity).

I agree that the human no longer gains expertise, but I am not sure that will matter in the long run.

The end result is a total and complete reliance on AI providers. It's not really a secret; that's why I started the article with the direct quote from Altman. They would love to see a world where nobody is able to do any type of development work without an LLM subscription enabled...even though they also admit that these models perform best when they are steered by someone who is highly capable and knowledgeable in the first place. They're powerful, but they are also VERY fallible.

Local AI would thwart this vision (which is why they're desperately also trying to ruin the hardware market at the same time), but the effort it would take to get local models to run as well as these frontier models is an odd thing to pursue just to avoid learning the fundamentals and gaining direct experience.

yes, this seems sadly very true for me, too.

In some ways, it feel like I'd be "stuck" at my 2024 capabilities if I didn't push myself constantly, and I'm really glad for every hour I spent learning/applying new stuff before that, because now it feels way harder.