Andrew Critch’s “greedy algorithm”: Whenever you catch yourself really wanting to do something you want to want, immediately reward yourself—by feeding yourself an M&M, or if that’s too difficult, immediately pumping your fist and saying “Yes!”
I have been doing this deliberately for a few months because I was starting to get fed up with fighting my instincts every time I chose to program for an hour and wanted to spend that hour reading science fiction, so I actually started standing and exercising to watch anime or read books I did not really need to, while rewarding myself for getting something done that was relatively high up my priority list by sitting / eating food I liked.
It seems like no matter how I try to rebalance the incentives and consequences, negative stimuli always seem to have a much stronger effect on my behaviour however, so although I get more studying done now I think it comes more from the significant increases in my perceived free time than from reward mechanisms.
It seems like no matter how I try to rebalance the incentives and consequences, negative stimuli always seem to have a much stronger effect on my behaviour
What are the negative stimuli? Have you looked into simple behaviorist methods for making studying less painful, instead of just making it more rewarding?
Assuming klfwip wants to maximize xyr own happiness, changing the situation by adding more pain wouldn’t help. It might increase the amount of studying, but klfwip could also do that by enjoying studying more (possibly by altering xyr study habits), which would have a greater expected utility because it also makes klfwip happier.
I study for perceived benefits that include happiness but are broad enough that I am willing to suffer in the short term for greater motivation. If someone put a gun to my head and ordered me to study, I would have to cooperate and probably be very productive, but I am just paranoid enough and value my current existence too much to let this happen.
However, after forcing myself to act in ways that violate my natural hyperbolic discounting for months, it seems to have sunk in a bit so it seems like even minor penalties for behavior I do not want to encourage are enough to change most of my habits if I am consistent enough.
I would not argue everyone should place themselves in self defined bootcamp to try and improve their abilities, but it has been an interesting experiment at least. Many organizations use similar tactics to brainwash members because it works, and it seems to be at least somewhat effective even when self administered.
The greatest negative stimuli of studying are hard to address for me, actual failure to comprehend a difficult set of problems for weeks is itself enough to make me want to give up completely at times, and the simplest ways to eliminate this would be by no longer caring about results or actually succeeding at everything I do. The first would remove most of my incentive to learn in the first place, the second I would love to do but I don’t expect this to be feasible any time in the near future.
There are probably effective ways to make studying more fun that I have not really explored though…
Studying in groups with other people dealing with the same problems seems to be effective but can be hard to do practically, Nicotine can be used to artificially associate actions with pleasant feelings, along with other drugs. If nicotine actually works as well as Gwern and some others suggest I may try it, but age and then financial constraints have been too limiting.
Andrew Critch’s “greedy algorithm”: Whenever you catch yourself really wanting to do something you want to want, immediately reward yourself—by feeding yourself an M&M, or if that’s too difficult, immediately pumping your fist and saying “Yes!”
I have been doing this deliberately for a few months because I was starting to get fed up with fighting my instincts every time I chose to program for an hour and wanted to spend that hour reading science fiction, so I actually started standing and exercising to watch anime or read books I did not really need to, while rewarding myself for getting something done that was relatively high up my priority list by sitting / eating food I liked.
It seems like no matter how I try to rebalance the incentives and consequences, negative stimuli always seem to have a much stronger effect on my behaviour however, so although I get more studying done now I think it comes more from the significant increases in my perceived free time than from reward mechanisms.
What are the negative stimuli? Have you looked into simple behaviorist methods for making studying less painful, instead of just making it more rewarding?
Or making the alternatives more painful?
Nope.
beeminder.com works for me…
You can try it first :) Tell me if it is worth it.
Assuming klfwip wants to maximize xyr own happiness, changing the situation by adding more pain wouldn’t help. It might increase the amount of studying, but klfwip could also do that by enjoying studying more (possibly by altering xyr study habits), which would have a greater expected utility because it also makes klfwip happier.
I study for perceived benefits that include happiness but are broad enough that I am willing to suffer in the short term for greater motivation. If someone put a gun to my head and ordered me to study, I would have to cooperate and probably be very productive, but I am just paranoid enough and value my current existence too much to let this happen.
However, after forcing myself to act in ways that violate my natural hyperbolic discounting for months, it seems to have sunk in a bit so it seems like even minor penalties for behavior I do not want to encourage are enough to change most of my habits if I am consistent enough.
I would not argue everyone should place themselves in self defined bootcamp to try and improve their abilities, but it has been an interesting experiment at least. Many organizations use similar tactics to brainwash members because it works, and it seems to be at least somewhat effective even when self administered.
The greatest negative stimuli of studying are hard to address for me, actual failure to comprehend a difficult set of problems for weeks is itself enough to make me want to give up completely at times, and the simplest ways to eliminate this would be by no longer caring about results or actually succeeding at everything I do. The first would remove most of my incentive to learn in the first place, the second I would love to do but I don’t expect this to be feasible any time in the near future.
There are probably effective ways to make studying more fun that I have not really explored though… Studying in groups with other people dealing with the same problems seems to be effective but can be hard to do practically, Nicotine can be used to artificially associate actions with pleasant feelings, along with other drugs. If nicotine actually works as well as Gwern and some others suggest I may try it, but age and then financial constraints have been too limiting.