I'd say watch out for the "One brain" fallacy/cognitive bias in all of these responses.
There are lots of factors (age, discipline, interest in what you're doing, etc) but I think a lot of people overlook that one big component of this is brain chemistry. Some athletes are endurance athletes, some athletes are sprinters. It seems obvious that you wouldn't expect Usain Bolt to run the fastest marathon just because he's a runner.
Our brains are just as genetically varied as other physical characteristics. For me, I know that my peak productivity is from 11 pm - 4 am and it goes down exponentially if I have to do non-fun tasks or if I'm in a stressful crunch as opposed to a positive crunch. I also know that after 3-4 months of 10-12 hours, I need to go back down to 5-6 hours until I get excited again. And for some strange reason, if I don't feel something is required, I can leisurely work on something for 12-16 hours a day indefinitely (researching, reading conference papers, learning). If I have to debug code nasty bugs like multithreaded data races, I can't be productive at those lengths of time so I try to mix fun work with this type of dreaded work.
Know what works for you and build your work schedule around that. That's how we work at our startup.
There are lots of factors (age, discipline, interest in what you're doing, etc) but I think a lot of people overlook that one big component of this is brain chemistry. Some athletes are endurance athletes, some athletes are sprinters. It seems obvious that you wouldn't expect Usain Bolt to run the fastest marathon just because he's a runner.
Our brains are just as genetically varied as other physical characteristics. For me, I know that my peak productivity is from 11 pm - 4 am and it goes down exponentially if I have to do non-fun tasks or if I'm in a stressful crunch as opposed to a positive crunch. I also know that after 3-4 months of 10-12 hours, I need to go back down to 5-6 hours until I get excited again. And for some strange reason, if I don't feel something is required, I can leisurely work on something for 12-16 hours a day indefinitely (researching, reading conference papers, learning). If I have to debug code nasty bugs like multithreaded data races, I can't be productive at those lengths of time so I try to mix fun work with this type of dreaded work.
Know what works for you and build your work schedule around that. That's how we work at our startup.