I'd also be very suspicious of this Pareto distribution, to me it implies that the 20% of work we're talking about is the hard part. Spitting out code without understanding it definitely sounds like the easy part, the part that doesn't require much if any thinking. I'd be much more interested to see a breakdown of TIME, not volume of code; how much time does using an LLM save (or not) in the context of a given task?
Whatever time is be saved (if?) will be taken up by the higher expectation of the worker. Imagine a manager’s monologue “now that you have AI, wouldnt’t it be easier to implement those features that we had to scrap because we were understaffed? Soon we’ll be even more understaffed and you’ll have to work harder.”
