Thanks to the popularity of generative AI, there has been an explosion of AI based coding assistants.
Then there is the recent McKinsey report on measuring developer productivity. Management has again fallen back into the trap of measuring dev outputs (rather than outcomes). They feel that somehow, if they can improve dev productivity numbers using coding assistants, they can score brownie points over their peers and impress their bosses.
Software engineering is usually perceived as the act of writing code. We as software engineers also obsess a lot about the code aspect of it. In reality, a large chunk of a software engineer’s time is actually spent talking to other people, understanding the problem, thinking of tradeoffs and coming up with a solution in your head or a piece of paper. Coding is just the act of putting all these things into code. An AI coding assistant may help you with the code, but it won’t help you with understanding the problem. Using a coding assistant may not move the needle as much as the management thinks it would.