For some reason you want to do them. It's not like some dummy algorithm to decide what you do right now. What you want is a very complex computation which is ba

For some reason you want to do them. It's not like some dummy algorithm to decide what you do right now. What you want is a very complex computation which is basically already done and takes into account everything you know. There's no reason to try to outsmart yourself. If you are excited about the future associated with some task (or about the future in general) you are excited about doing the task. Some grim outlook on the future makes it rational to optimize short term results. E.g. sleep deprivation gives a glimpse of that. Basically seems pointless to fight with yourself. Maybe you don't really want the long-term task + result package. Or maybe wishful thinking is causing you to fantasize about getting a result without doing the task. (I'm sorry about using second person, seems most straightforward way to communicate it). I guess there's also some element of narration creating fictional character - e.g. I can't stop playing chess - playing chess is who I am. Which basically is daydreaming again. And btw it all sound pretty analytical but in practice is mostly value driven. But if necessary, I will debug the stack deeply. Honesty is required. It goes deep and into intimate places. I would leave my wife and kids if I wouldn't want to be with them (and I think that's best for everybody). Given that I do and want them to be happy a lot of stuff follows effortlessly. I think many people are afraid to question these deep assumptions. But it's worth doing it. Otherwise it's castles on sand. Back when I was starting with debugging I once deeply considered our marriage since I needed to take some trash out :))