Cached Thoughts

One of the sin­gle great­est puz­zles about the hu­man brain is how the damn thing works at all when most neu­rons fire 10–20 times per sec­ond, or 200Hz tops. In neu­rol­ogy, the “hun­dred-step rule” is that any pos­tu­lated op­er­a­tion has to com­plete in at most 100 se­quen­tial steps—you can be as par­allel as you like, but you can’t pos­tu­late more than 100 (prefer­ably fewer) neu­ral spikes one af­ter the other.

Can you imag­ine hav­ing to pro­gram us­ing 100Hz CPUs, no mat­ter how many of them you had? You’d also need a hun­dred billion pro­ces­sors just to get any­thing done in re­al­time.

If you did need to write re­al­time pro­grams for a hun­dred billion 100Hz pro­ces­sors, one trick you’d use as heav­ily as pos­si­ble is caching. That’s when you store the re­sults of pre­vi­ous op­er­a­tions and look them up next time, in­stead of re­com­put­ing them from scratch. And it’s a very neu­ral idiom—recog­ni­tion, as­so­ci­a­tion, com­plet­ing the pat­tern.

It’s a good guess that the ac­tual ma­jor­ity of hu­man cog­ni­tion con­sists of cache lookups.

This thought does tend to go through my mind at cer­tain times.

There was a won­der­fully illus­tra­tive story which I thought I had book­marked, but couldn’t re-find: it was the story of a man whose know-it-all neigh­bor had once claimed in pass­ing that the best way to re­move a chim­ney from your house was to knock out the fire­place, wait for the bricks to drop down one level, knock out those bricks, and re­peat un­til the chim­ney was gone. Years later, when the man wanted to re­move his own chim­ney, this cached thought was lurk­ing, wait­ing to pounce . . .

As the man noted af­ter­ward—you can guess it didn’t go well—his neigh­bor was not par­tic­u­larly knowl­edge­able in these mat­ters, not a trusted source. If he’d ques­tioned the idea, he prob­a­bly would have re­al­ized it was a poor one. Some cache hits we’d be bet­ter off re­com­put­ing. But the brain com­pletes the pat­tern au­to­mat­i­cally—and if you don’t con­sciously re­al­ize the pat­tern needs cor­rec­tion, you’ll be left with a com­pleted pat­tern.

I sus­pect that if the thought had oc­curred to the man him­self—if he’d per­son­ally had this bright idea for how to re­move a chim­ney—he would have ex­am­ined the idea more crit­i­cally. But if some­one else has already thought an idea through, you can save on com­put­ing power by caching their con­clu­sion—right?

In mod­ern civ­i­liza­tion par­tic­u­larly, no one can think fast enough to think their own thoughts. If I’d been aban­doned in the woods as an in­fant, raised by wolves or silent robots, I would scarcely be rec­og­niz­able as hu­man. No one can think fast enough to re­ca­pitu­late the wis­dom of a hunter-gath­erer tribe in one life­time, start­ing from scratch. As for the wis­dom of a liter­ate civ­i­liza­tion, for­get it.

But the flip side of this is that I con­tinu­ally see peo­ple who as­pire to crit­i­cal think­ing, re­peat­ing back cached thoughts which were not in­vented by crit­i­cal thinkers.

A good ex­am­ple is the skep­tic who con­cedes, “Well, you can’t prove or dis­prove a re­li­gion by fac­tual ev­i­dence.” As I have pointed out el­se­where,1 this is sim­ply false as prob­a­bil­ity the­ory. And it is also sim­ply false rel­a­tive to the real psy­chol­ogy of re­li­gion—a few cen­turies ago, say­ing this would have got­ten you burned at the stake. A mother whose daugh­ter has can­cer prays, “God, please heal my daugh­ter,” not, “Dear God, I know that re­li­gions are not al­lowed to have any falsifi­able con­se­quences, which means that you can’t pos­si­bly heal my daugh­ter, so . . . well, ba­si­cally, I’m pray­ing to make my­self feel bet­ter, in­stead of do­ing some­thing that could ac­tu­ally help my daugh­ter.”

But peo­ple read “You can’t prove or dis­prove a re­li­gion by fac­tual ev­i­dence,” and then, the next time they see a piece of ev­i­dence dis­prov­ing a re­li­gion, their brain com­pletes the pat­tern. Even some athe­ists re­peat this ab­sur­dity with­out hes­i­ta­tion. If they’d thought of the idea them­selves, rather than hear­ing it from some­one else, they would have been more skep­ti­cal.

Death. Com­plete the pat­tern: “Death gives mean­ing to life.”

It’s frus­trat­ing, talk­ing to good and de­cent folk—peo­ple who would never in a thou­sand years spon­ta­neously think of wiping out the hu­man species—rais­ing the topic of ex­is­ten­tial risk, and hear­ing them say, “Well, maybe the hu­man species doesn’t de­serve to sur­vive.” They would never in a thou­sand years shoot their own child, who is a part of the hu­man species, but the brain com­pletes the pat­tern.

What pat­terns are be­ing com­pleted, in­side your mind, that you never chose to be there?

Ra­tion­al­ity. Com­plete the pat­tern: “Love isn’t ra­tio­nal.”

If this idea had sud­denly oc­curred to you per­son­ally, as an en­tirely new thought, how would you ex­am­ine it crit­i­cally? I know what I would say, but what would you? It can be hard to see with fresh eyes. Try to keep your mind from com­plet­ing the pat­tern in the stan­dard, un­sur­pris­ing, already-known way. It may be that there is no bet­ter an­swer than the stan­dard one, but you can’t think about the an­swer un­til you can stop your brain from filling in the an­swer au­to­mat­i­cally.

Now that you’ve read this, the next time you hear some­one un­hesi­tat­ingly re­peat­ing a meme you think is silly or false, you’ll think, “Cached thoughts.” My be­lief is now there in your mind, wait­ing to com­plete the pat­tern. But is it true? Don’t let your mind com­plete the pat­tern! Think!

1See ’Reli­gion’s Claim to be Non-Disprov­able,” in Map and Ter­ri­tory.