I have found over the decades that my debugging ability declines as fatigue increases. So if the goal is working code, there is a point where the additional work leads to delay whenever something is broken or has to be changes. Hack attacks of short duration is great to get some code off the ground, but anything that must be maintained must be clear enough to be revisited later. I have tried to average 45-50 hours/wk max, although sometimes out-of-work swaps for in-work time.

Wow, that's a wishy-washy article. Is it "AI"-written, and human-edited?

How long people actually work is not the same as how long they ought to.

And that's the source of controversy, with supervisors almost universally coming done on the side of "ought to work a lot more", or at least do things that supervisors perceive to be working a lot more.

They are getting paid an average of 42 hoursba week. The vast majority of people I know (especially in tech) are actually working less hours than this.

I remember reading an article about British workers during WW2 who were willing to work any number of hours to beat the Nazis. Initially their productivity spiked as they worked more hours. Eventually it crashed, and they did a study to identify the most productive number of hours. 35.

I remember another article from a Harvard business review where they studied the number of hours worked which was correlated with the highest promotion rate in business. 45.

So, if you’re riveting airplanes, or other manual work, 35 hours per week seem to be maximal. For office work, 45 hours per week at least gets you the best results from the business noticing your effort. I shoot for 45 hours a week on average, but sometimes go as high as 80, when sprinting, and as low as 25, when trying to recover from those sprints. Seems to have worked well for my career and work/life balance.