Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio.
This is: My Algorithm for Beating Procrastination, published by lukeprog on the AI Alignment Forum.
Part of the sequence: The Science of Winning at Life
After three months of practice, I now use a single algorithm to beat procrastination most of the times I face it.1 It probably won't work for you quite like it did for me, but it's the best advice on motivation I've got, and it's a major reason I'm known for having the "gets shit done" property. There are reasons to hope that we can eventually break the chain of akrasia; maybe this post is one baby step in the right direction.
How to Beat Procrastination explained our best current general theory of procrastination, called "temporal motivation theory" (TMT). As an exercise in practical advice backed by deep theories, this post explains the process I use to beat procrastination — a process implied by TMT.
As a reminder, here's a rough sketch of how motivation works according to TMT:
Or, as Piers Steel summarizes:
Decrease the certainty or the size of a task's reward — its expectancy or its value — and you are unlikely to pursue its completion with any vigor. Increase the delay for the task's reward and our susceptibility to delay — impulsiveness — and motivation also dips.
Of course, my motivation system is more complex than that. P.J. Eby likens TMT (as a guide for beating procrastination) to the "fuel, air, ignition, and compression" plan for starting your car: it might be true, but a more useful theory would include details and mechanism.
That's a fair criticism. Just as an fMRI captures the "big picture" of brain function at low resolution, TMT captures the big picture of motivation. This big picture helps us see where we need to work at the gears-and-circuits level, so we can become the goal-directed consequentialists we'd like to be.
So, I'll share my four-step algorithm below, and tackle the gears-and-circuits level in later posts.
Step 1: Notice I'm procrastinating.
This part's easy. I know I should do the task, but I feel averse to doing it, or I just don't feel motivated enough to care. So I put it off, even though my prefrontal cortex keeps telling me I'll be better off if I do it now. When this happens, I proceed to step 2.
Step 2: Guess which unattacked part of the equation is causing me the most trouble.
Now I get to play detective. Which part of the equation is causing me trouble, here? Does the task have low value because it's boring or painful or too difficult, or because the reward isn't that great? Do I doubt that completing the task will pay off? Would I have to wait a long time for my reward if I succeeded? Am I particularly impatient or impulsive, either now or in general? Which part of this problem do I need to attack?
Actually, I lied. I like to play army sniper. I stare down my telescopic sight at the terms in the equation and interrogate them. "Is it you, Delay? Huh, motherfucker? Is it you? I've shot you before; don't think I won't do it again!"
But not everyone was raised on violent videogames. You may prefer a different role-play.
Anyway, I try to figure out where the main problem is. Here are some of the signs I look for:
When I imagine myself doing the task, do I see myself bored and distracted instead of engaged and interested? Is the task uncomfortable, onerous, or painful? Am I nervous about the task, or afraid of what might happen if I undertake it? Has the task's payoff lost its value to me? Perhaps it never had much value to me in the first place? If my answer to any of these questions is "Yes," I'm probably facing the motivation problem of low value.
Do I think I'm likely to succeed at the task? Do I think it's within my capabilities? Do I think I'll actually get the reward if I do succeed? If my answer to any of these questions is "No," I'm probably facing the problem of low expectancy.
Ho...
view more