“Code Was Never the Hard Part” Rewrites the History of Programming

Senko Rašić asks a series of wonderfully uncomfortable questions about the increasingly common claim that AI can automate coding because “code was never the hard part.”

If coding was easy, why were good programmers so difficult to hire?

Why the salaries, brutal interviews, burnout and decades spent learning the craft?

Why do we have The Art of Computer Programming, Structure and Interpretation of Computer Programs, The Pragmatic Programmer and entire university degrees devoted to something that was supposedly the easy part?

And my favorite:

If coding is easy, why is software so damn buggy?

Rašić makes the same challenge from the other direction. If figuring out what customers want was always the genuinely difficult part, why weren't product managers, usability researchers, business analysts and customer-success teams historically treated as the most valuable people in software companies?

The answer is probably that this was always a false choice.

Understanding why software should exist is hard.

Understanding how to build it well is also hard.

AI may dramatically reduce the cost of producing code, but that does not retroactively make implementation trivial. It changes which parts of the craft require scarce human effort.

Rašić's conclusion is the part worth keeping: developers should deepen their understanding of systems while also learning more about users, products and businesses. And whatever changes, we should not outsource our understanding, judgment, empathy or taste to AI.

That feels like a much healthier response to AI than pretending either that programming was never difficult or that nothing about programming is going to change.

Read “‘Code was never the hard part’ is an insult to all programmers.”