Good points in the article, I've been questioning this for a couple years now as AI makes everyone think we can solve everything. AI doesn't know how to create value.

The article talks about business value, but there is also technical value, and those two are tied together. AI doesn't understand either, but not many talk about technical value, i.e. a good working codebase, it's properly abstracted, easily extensible, reliable and robust.

> Okay, but the question that bugs me is why are so many people obsessed with increasing engineering speed?

The reason could be that technical value has been undervalued and ignored across the software industry. Engineering has been reduced to speed of delivery, in part because now every software is constantly updated and patched.

Another related reason is the overloading of the role of engineer. A couple decades ago it was a role that was mostly focused on technical problems and has increasingly become an extension of the business, solving business problems, in part because technical value has been ignored so much.

Now that AI will saturate software in so many ways, it's probably a good idea to focus again not just on business value but also technical value, which go hand-in-hand in most cases.