Working hurts less than procrastinating, we fear the twinge of starting

When you pro­cras­ti­nate, you’re prob­a­bly not pro­cras­ti­nat­ing be­cause of the pain of work­ing.

How do I know this? Be­cause on a mo­ment-to-mo­ment ba­sis, be­ing in the mid­dle of do­ing the work is usu­ally less painful than be­ing in the mid­dle of pro­cras­ti­nat­ing.

(Bolded be­cause it’s true, im­por­tant, and nearly im­pos­si­ble to get your brain to re­mem­ber—even though a few mo­ments of re­flec­tion should con­vince you that it’s true.)

So what is our brain flinch­ing away from, if not the pain of do­ing the work?

I think it’s flinch­ing away from the pain of the de­ci­sion to do the work—the mo­men­tary, im­me­di­ate pain of (1) dis­en­gag­ing your­self from the (prob­a­bly very small) flow of re­in­force­ment that you’re get­ting from read­ing a ran­dom unim­por­tant In­ter­net ar­ti­cle, and (2) pay­ing the en­ergy cost for a pre­frontal over­ride to ex­ert con­trol of your own be­hav­ior and be­gin work­ing.

Thanks to hy­per­bolic dis­count­ing (i.e., weight­ing val­ues in in­verse pro­por­tion to their tem­po­ral dis­tance) the in­stant pain of dis­en­gag­ing from an In­ter­net ar­ti­cle and pay­ing a pre­frontal over­ride cost, can out­weigh the slightly more dis­tant (min­utes in the fu­ture, rather than sec­onds) pain of con­tin­u­ing to pro­cras­ti­nate, which is, once again, usu­ally more painful than be­ing in the mid­dle of do­ing the work.

I think that hy­per­bolic dis­count­ing is far more ubiquitous as a failure mode than I once re­al­ized, be­cause it’s not just for com­men­su­rate-seem­ing trade­offs like smok­ing a cigarette in a minute ver­sus dy­ing of lung can­cer later.

When it comes to pro­cras­ti­nat­ing, the ob­vi­ous, salient, com­men­su­rate-seem­ing trade­off, is be­tween the (as­sumed) plea­sure of read­ing a ran­dom In­ter­net ar­ti­cle now, ver­sus the (as­sumed) pain of do­ing the work now. But this, as I said above, is not where I think the real trade­off is; events that are five min­utes away are too dis­tant to dom­i­nate the thought pro­cess of a hy­per­bolic dis­counter like a hu­man. In­stead our thought pro­cesses are dom­i­nated by the prospec­tive im­me­di­ate pain of a thought, a cost that isn’t even salient as some­thing to be traded off. “Work­ing” is an ob­vi­ous, salient event, and “read­ing ran­dom ar­ti­cles” seems like an event. But “pay­ing a small twinge of pain to make the de­ci­sion to stop pro­cras­ti­nat­ing now, ex­ert­ing a bit of frontal over­ride, and not get­ting to read the next para­graph of this ran­dom ar­ti­cle” is so map-level that we don’t even fo­cus on it as a ma­nipu­la­ble ter­ri­tory, a cost to be traded off; it is a trans­par­ent thought.

The real dam­age done by hy­per­bolic dis­count­ing is for thoughts that are only very slightly painful, and yet, these slight pains be­ing im­me­di­ate, they man­age to dom­i­nate ev­ery­thing else in our calcu­la­tion. And be­ing trans­par­ent, we aren’t even aware that’s what’s hap­pen­ing. “Be­ware of im­me­di­ately triv­ially painful trans­par­ent thoughts”, one might say.

Similarly, you may read a mediocre book for an hour, in­stead of a good book, be­cause if you first spent a few min­utes to search your library to ob­tain a bet­ter book, that would be an im­me­di­ate cost—not that search­ing your library is all that un­pleas­ant, but you’d have to pay an im­me­di­ate ac­ti­va­tion cost to do that in­stead of tak­ing the path of least re­sis­tance and grab­bing the first thing in front of you. It’s a hy­per­bol­i­cally dis­counted trade­off that you make with­out re­al­iz­ing it, be­cause the cost you’re re­fus­ing to pay isn’t com­men­su­rate enough with the pay­off you’re for­go­ing to be salient as an ex­plicit trade­off.

A re­lated note that I might as well dump into this post: I’m start­ing to think that pro­cras­ti­na­tion by read­ing ran­dom ar­ti­cles does not cause you to rest, that is, you do not re­gain men­tal en­ergy from it. Suc­cess and hap­piness cause you to re­gain willpower; what you need to heal your mind from any dam­age sus­tained by work­ing is not in­ac­tivity, but re­li­ably solv­able prob­lems which re­li­ably de­liver ex­pe­rienced jolts of pos­i­tive re­in­force­ment. Put­ting in the effort to read a good book may do this; play­ing a good com­puter game may do this; read­ing ran­dom In­ter­net ar­ti­cles, or play­ing bad games, prob­a­bly won’t. Literal men­tal ex­haus­tion might mean that you don’t have enough en­ergy left to read a good book—or that you don’t have enough en­ergy left to pay the im­me­di­ate cost of search­ing your library for good read­ing ma­te­rial in­stead of mediocre read­ing ma­te­rial—but in this case you shouldn’t be read­ing ran­dom on­line ar­ti­cles. You should be sit­ting with your eyes closed listen­ing to mu­sic, or pos­si­bly even nap­ping; if deal­ing with a truly ex­hausted brain, read­ing ran­dom ar­ti­cles is prob­a­bly too much effort.

If you don’t feel good while read­ing a lot of for­get­table on­line ar­ti­cles, and you don’t feel re­newed af­ter do­ing so, your in­tu­itive the­ory which says that this is how to rest is mis­taken, and you need to look for other ways to rest in­stead—more ac­tive ways to re­gain willpower, less ac­tive ways to re­cover from im­me­di­ate ex­haus­tion. In gen­eral, poor perfor­mance of­ten in­di­cates poor mod­els; if some­thing seems in­cred­ibly difficult to pre­dict or ma­nipu­late, it may be that you have mis­taken be­liefs about it, in­clud­ing trans­par­ent mis­takes that are non­ques­tioned be­cause they are non­salient. This in­cludes poor perfor­mance on the prob­lem of rest­ing.

Hope­fully pub­lish­ing this post will help me live up to it.