[Link] Why Generalists Triumph in a Specialized World


Sum­mary of why this is a worth­while read for peo­ple in­ter­ested in EA:

  • Read­ing this book helped me see the benefit of breadth for solv­ing par­tic­u­larly “wicked” en­vi­ron­ments (‘Spe­cial­ists flour­ish in such “kind” learn­ing en­vi­ron­ments, where pat­terns re­cur and feed­back is quick and ac­cu­rate. By con­trast, gen­er­al­ists flour­ish in “wicked” learn­ing en­vi­ron­ments, where pat­terns are harder to dis­cern and feed­back is de­layed and/​or in­ac­cu­rate.’)

  • It seems to me that many of the prob­lems that effec­tive al­tru­ists are tack­ling are in “wicked” en­vi­ron­ments, with de­layed feed­back, difficult to dis­cern pat­terns. (“Golf, chess, clas­si­cal-mu­sic perfor­mance, fire­fight­ing and anes­the­siol­ogy are (in this sense) kind learn­ing en­vi­ron­ments; ten­nis and jazz are less kind; emer­gency-room medicine, tech­nolog­i­cal in­no­va­tion and geopoli­ti­cal fore­cast­ing are down­right wicked — as is much of the rest of mod­ern life.”)

  • There­fore, the im­pli­ca­tion of the book is that many peo­ple solv­ing prob­lems in EA would benefit from be­ing gen­er­al­ists. I sus­pect that this is already of­ten the case. I think this book is use­ful be­cause it pro­vides a case for why that is benefi­cial, which I be­lieve could lead to more effec­tive prob­lem solv­ing on some of the thorny EA and EA ad­ja­cent prob­lems.

  • The down­sides of the book are that it has some ap­par­ent in­con­sis­ten­cies and that it has many sto­ries and less con­crete data. I think that’s the na­ture of the topic, and I still think it’s worth read­ing – at least the ar­ti­cle, and likely the book – for peo­ple in EA that are try­ing to tackle par­tic­u­larly com­plex prob­lems.