What are your thoughts on my career change options? (AI public policy)

I am go­ing through a pe­riod of tran­si­tion in my own life and am go­ing to change ca­reer. Whilst I think about these things I am do­ing a 6 month fel­low­ship at Newspeak House—if you’re in Lon­don check the events cal­en­dar or mes­sage me and swing by some time.

80000 Hours coach­ing advice

I did the 80000 hours coach­ing call this week. so I thought I’d ask for your thoughts on what was said:

  • I would re­ally recom­mend sign­ing up to 80k coach­ing. Also set a flag so you don’t miss the email.

  • I have a 2:1 from ~7th best Uni in the coun­try. The coach reck­oned get­ting on to a world top 10 ma­chine learn­ing Phd would be difficult, so it prob­a­bly wouldn’t be worth my time ap­ply­ing. I will prob­a­bly fol­low this ad­vice, though I don’t un­der­stand how it can both be true that AI ethics can be hugely tal­ent con­strained whilst also not ac­cept­ing good can­di­dates. (Per­haps the in­dus­try has high stan­dards which in­ter­ested EAs have to meet, so it’s not the in­dus­try which is tal­ent con­strained but EAs? Like­wise, per­haps I’m not a good can­di­date)

  • Their ad­vice was to try and go into pub­lic policy po­si­tions in UK/​EA in AI or try to be­come an MP. Seems like good ad­vice. I have a list of peo­ple to talk to and ac­tions to do.

A bit about Newspeak House

Fol­low­ing the call, I chat­ted to Ed, the Dean of Newspeak House. First a word about the col­lege. “Newspeak House is in­de­pen­dent res­i­den­tial col­lege founded in 2015 to study, nur­ture and in­spire emerg­ing com­mu­ni­ties of prac­tice across UK pub­lic sec­tor and civil so­ciety.” This is in my ex­pe­rience what it is at­tempt­ing to do, though its meth­ods are non-stan­dard. There are 7 live-in Fel­lows with with jobs and lives who study and teach trans­fer­able skills to the com­mu­ni­ties which use Newspeak House to host their events (by virtue of liv­ing in the space).

As an aside, I tried to write a piece eval­u­at­ing the im­pact of this, but it’s hard. I guess I will write an 80000 hours style ques­tion­naire and send it to Fel­lows/​com­mu­nity or­ganisers who use our space and see if there are trends in terms of im­pact. Any ad­vice wel­come, I will likely write some­thing at some stage, but if I am go­ing to make a lot of effort I’d like to not guess­ing so wildly as for it to be non­sense..

Ed, Newspeak House Dean’s, advice

Ed is of the opinion that us­ing sys­tems to your ad­van­tage you can have a greater im­pact. Rather than em­ploy 1 re­searcher, for the same cost, Newspeak House has 7 fel­lows who provide a much wider range of thoughts and con­nec­tions. While talk­ing about my fu­ture, he was of the opinion that to meet my own goals, I should:

  • Get fund­ing ei­ther through free­lance work or EA grants

  • Work for free in the AI pub­lic policy space out­side of an institution

  • At­tend all of the AI events in par­li­a­ment (which are sup­pos­edly open to the pub­lic—more re­search needed)

  • Create net­works with the movers and shak­ers, con­nect peo­ple where ad­van­ta­geous, talk to those in par­li­a­ment about EA ideas

  • Do my own re­search, tell re­searchers if I find any­thing in­ter­est­ing and let them pub­lish.

I don’t know how the im­pact of this route would com­pare to a more stan­dard Civil Ser­vice fast stream path. I am open to the idea it could be more effec­tive—hav­ing lived at NH for a cou­ple of weeks now, I have already met loads of well con­nected in­di­vi­d­u­als and seen Ed cre­ate events just but get­ting two peo­ple to talk. Some­one has to do the work, but some­one of­ten already is and if you can con­nect 10x as many peo­ple you can in the­ory do 10x as much good. That said, it is hard (as above) to mea­sure im­pact and per­haps it is valuable to re­ally be in the room when de­ci­sions are be­ing made.

Emo­tion­ally speak­ing I am ini­tially skep­ti­cal of Ed’s idea, mainly be­cause I would like a more sta­ble job, but my ca­reer is the area of my life where I would like to make re­ally effi­cient choices, so I’d like to hear your thoughts.

Sum up

I would like to work for the max­i­mal good of con­scious crea­tures. Seems like AI policy is a good way to go, but it’s un­clear if a stan­dard or sys­tems-based ap­proach will be more effec­tive. I’d love to hear your thoughts and in­tend to change my ap­proach based on them.