What is up with carbon dioxide and cognition? An offer

One or two re­search groups have pub­lished work on car­bon diox­ide and cog­ni­tion. The state of the pub­lished liter­a­ture is con­fus­ing.

Here is one pa­per on the topic. The au­thors in­ves­ti­gate a pro­pri­etary cog­ni­tive bench­mark, and ex­per­i­men­tally ma­nipu­late car­bon diox­ide lev­els (with­out af­fect­ing other mea­sures of air qual­ity). They find im­plau­si­bly large effects from in­creased car­bon diox­ide con­cen­tra­tions.

If the re­ported effects are real and the sug­gested in­ter­pre­ta­tion is cor­rect, I think it would be a big deal. To put this in per­spec­tive, car­bon diox­ide con­cen­tra­tions in my room vary be­tween 500 and 1500 ppm de­pend­ing on whether I open the win­dows. The ex­per­i­ment re­ports on cog­ni­tive effects for mov­ing from 600 and 1000 ppm, and finds sig­nifi­cant effects com­pared to in­terindi­vi­d­ual differ­ences.

I haven’t spent much time look­ing into this (maybe 30 min­utes, and an­other 30 min­utes to write this post). I ex­pect that if we spent some time look­ing into in­door CO2 we could have a much bet­ter sense of what was go­ing on, by some com­bi­na­tion of bet­ter liter­a­ture re­view, dis­cus­sion with ex­perts, look­ing into the bench­mark they used, and just gen­er­ally think­ing about it.

So, here’s a pro­posal:

  • If some­one looks into this and writes a post that im­proves our col­lec­tive un­der­stand­ing of the is­sue, I will be will­ing to buy part of an as­so­ci­ated cer­tifi­cate of im­pact, at a price of around $100*N, where N is my own to­tally made up es­ti­mate of how many hours of my own time it would take to pro­duce a similarly use­ful writeup. I’d buy up to 50% of the cer­tifi­cate at that price.

  • Whether or not they want to sell me some of the cer­tifi­cate, on May 1 I’ll give a $500 prize to the au­thor of the best pub­li­cly-available anal­y­sis of the is­sue. If the best anal­y­sis draws heav­ily on some­one else’s work, I’ll use my dis­cre­tion: I may split the prize ar­bi­trar­ily, and may give it to the ear­lier post even if it is not quite as ex­cel­lent.

Some clar­ifi­ca­tions:

  • The met­ric for qual­ity is “how use­ful it is to Paul.” I hope that’s a use­ful proxy for how use­ful it is in gen­eral, but no guaran­tees. I am gen­er­ally a pretty skep­ti­cal per­son. I would care a lot about even a mod­est but well-es­tab­lished effect on perfor­mance.

  • Th­ese don’t need to be new analy­ses, ei­ther for the prize or the pur­chase.

  • I re­serve the right to re­solve all am­bi­gui­ties ar­bi­trar­ily, and in the end to do what­ever I feel like. But I promise I am gen­er­ally a nice guy.

  • I posted this 2 weeks ago on the EA fo­rum and haven’t had se­ri­ous tak­ers yet.

(Thanks to An­drew Critch for men­tion­ing these re­sults to me and Jes­sica Tay­lor for lend­ing me a CO2 mon­i­tor so that I could see vari­abil­ity in in­door CO2 lev­els. I apol­o­gize for de­liber­ately not do­ing my home­work on this post.)