I think its a personal call. If as an engineer you value your art you will go deep, be the best at what you do, and be valuable in the correct context. If you're a product manager you just need to know enough to communicate with engineers or to make informed technology decisions. Either way, whatever you are or do, there is in my opinion more merit in pursing depth in whatever you do.

Being a quick learner is a good thing, but there comes a point where thorough knowledge and broad overviews are required to take you to the next level.

Learning new skills is always a gamble, what's valuable today may not be much use tomorrow. But still, the more tools in your toolbox, the wider the variety of problems you can solve. HR departments (and now AI) are certainly over-obsessed with seeing acronyms on your resume, but there's a basic truth behind that.

I think there is not really a contradiction between having expertise and giving value but if you feel there is it is because of politics, not in the left/right sense but office politics.

That is, the bosses son isn’t going to feel there is a contradiction here but you’re going to feel it when they try to explain to you why he got a promotion and he didn’t.

Now some places are healthier than others and some are more fair than others.

I think to get the whole picture you have to be able to hold a few different viewpoints about a situation. In the end you deliver value as part of a team and that can have many corollaries such as: prioritizing what works for the team as opposed to myself (so far as I get what I need and feel OK about it), getting some of my personal development through side projects, recognizing that my team puts limitations on how much value I can make and deciding how much I want to live with it, press for change in the organization or move on.

Specialization and deep knowledge is what allows you to command large salaries. Its supply and demand if you think about it, few people will put in the world to become domain experts. Steven Cleary comes to mind as he's a deep domain expert on .NET which millions of people can program in, but who is the deep expert in the field?

In the age of AI, expertise alone isn't going to cut. It has to be super deep expertise to outperform AI.

There’s generally a vast chasm between the expertise companies think they need, and what they actually utilise.

The problem is this expertise is what they evaluate at interview, and so if you want to remain marketable you need to keep up.

Maintaining and expanding that expertise then falls to you as a recreational project for your personal time, because you’ll get to a point where your day job teaches you nothing new.

You’re not wrong — you're just stuck in a system that rewards credentials, not contribution. We’re building a protocol to fix that.