The Unreasonable Effectiveness of Certain Questions

Cross-posted on my blog: http://​​gary­​​the-un­rea­son­able-effec­tive­ness-of-cer­tain-ques­tions/​​

About a year ago I was sit­ting around try­ing to grok the con­cept of Evil — where does it come from and how does it work? After a few hours of spin­ning in cir­cles, I ex­pe­rienced a sud­den shift. My mind con­jured up the ques­tion: “Is this a thing out in the world or just a pro­jec­tion?” (Map vs Ter­ri­tory). Im­me­di­ately, a part of my mind replied with “Well, this may not be any­thing other than a story we tell about the be­hav­ior of peo­ple we dis­like”. Let’s ig­nore the truth value for to­day and no­tice the pro­cess. I’m in­ter­ested in this mechanism of how a sim­ple query — check­ing if I’m look­ing at a con­fu­sion of map with the ter­ri­tory — was able to in­stantly re­frame a prob­lem in a way that al­lowed me to effortlessly make a men­tal leap. What’s fas­ci­nat­ing is that you don’t even need some­one else’s brain to come up with these ques­tions (al­though that of­ten helps) — you can try to ex­plain your prob­lem to a rub­ber duck which cre­ates a con­ver­sa­tion with your­self and gen­er­ates queries, or just go through a list of things to ask your­self when stuck.

There are a few differ­ent cat­e­gories of these types of queries and many ex­am­ples of each. For in­stance, when think­ing about plans we can ask our­selves to perform pre­hind­sight/​in­ner simu­la­tor or refer­ence class fore­cast­ing/​out­side view. When in­tro­spect­ing on our own be­hav­ior, we can perform sen­tence com­ple­tion to check for limit­ing be­liefs, ask ques­tions like “Why aren’t I done yet?” or “What can I do to 10x my re­sults?”. When think­ing about prob­lems or situ­a­tions, we can ask our­selves to in­vert, re­frame into some­thing falsifi­able, and taboo your words or perform parad­jitsu. Or con­sider the mir­a­cle ques­tion: Imag­ine you wake up and the prob­lem is en­tirely solved — what do you see, as con­cretely as pos­si­ble, such that you know this is true?

So “we know more than we can tell” — some­where in our head of­ten lies the an­swer, if only we could get to it. In some sense, parts of our brain are not speak­ing to each other (do they even share the same on­tolo­gies?) ex­cept through our lan­guage pro­ces­sor, and only then if the sen­tences are con­structed in spe­cific ways. This may make you feel re­lieved if you think you can rely on your sub­con­scious pro­cess­ing — which may have ac­cess to this knowl­edge — to guide you to effec­tive ac­tion, or ter­rified if you need to use con­scious rea­son­ing to think through a chain of con­se­quences.

My thoughts on Evil have con­tinued to evolve since that ini­tial rev­e­la­tion, par­tially driven by try­ing new queries on the con­cept (and par­tially from fi­nally read­ing Niet­zsche). Once you have a set of tools to throw at prob­lems, the bot­tle­neck to clearer think­ing be­comes re­mem­ber­ing to ap­ply them and ac­tu­ally hav­ing the time to do so. This makes me won­der about peo­ple that have formed habits to au­to­mat­i­cally ap­ply a litany of these men­tal moves when­ever ap­proach­ing a prob­lem — how much of their effec­tive­ness and in­tel­li­gence can this ex­plain?