AI coding tools are effective for many because, unfortunately, our work has become increasingly repetitive. When someone marvels at how a brief prompt can produce functioning code, it simply means the AI has delivered a more imaginative or elaborate specification than that person could envision, even if the resulting code is merely a variation of what has already been written countless times before. Maybe there's nothing wrong with that, as not everyone is fortunate enough to work on new problems and get to implement new ideas. It's just that repetitive work is bound to be automated away and therefore we will see the problems in Rich's rants.
That said, luminaries like Rob Pike and Rich Hickey do not have the above problem. They have the calibre and the freedom to push the boundaries, so to them the above problem is even amplified.
Personally I wish the IT industry can move forward to solve large-scale new problems, just like we did in the past 20 years: internet, mobile, the cloud, the machine learning... They created enormous opportunities (or the enormous opportunities of having software eat the world called for the opportunities?). I'm not sure we will be so lucky for the coming years, but we certainly should try.
That said, luminaries like Rob Pike and Rich Hickey do not have the above problem. They have the calibre and the freedom to push the boundaries, so to them the above problem is even amplified.
Personally I wish the IT industry can move forward to solve large-scale new problems, just like we did in the past 20 years: internet, mobile, the cloud, the machine learning... They created enormous opportunities (or the enormous opportunities of having software eat the world called for the opportunities?). I'm not sure we will be so lucky for the coming years, but we certainly should try.