Announcement: AI alignment prize round 3 winners and next round

We (Zvi Mow­show­itz and Vladimir Slep­nev) are happy to an­nounce the re­sults of the third round of the AI Align­ment Prize, funded by Paul Chris­ti­ano. From April 15 to June 30 we re­ceived en­tries from 12 par­ti­ci­pants, and are award­ing $10,000 to two win­ners.

We are also an­nounc­ing the fourth round of the prize, which will run un­til De­cem­ber 31 of this year un­der slightly differ­ent rules. More de­tails be­low.

The winners

First prize of $7,500 goes to Vanessa Kosoy for The Learn­ing-The­o­retic AI Align­ment Re­search Agenda. We feel this is much more ac­cessible than pre­vi­ous writ­ing on this topic, and gives a lot of promis­ing ideas for fu­ture re­search. Most im­por­tantly, it ex­plains why she is work­ing on the prob­lems she’s work­ing on, in con­crete enough ways to en­courage pro­duc­tive de­bate and dis­agree­ment.

Se­cond prize of $2,500 goes to Alexan­der Turner for the posts Wor­ry­ing About the Vase: Whitelist­ing and Over­com­ing Cling­i­ness in Im­pact Mea­sures. We are es­pe­cially happy with the amount of good dis­cus­sion these posts gen­er­ated.

We will con­tact each win­ner by email to ar­range trans­fer of money. Many thanks to ev­ery­one else who sent in their work!

The next round

We are now an­nounc­ing the fourth round of the AI Align­ment Prize. Due the drop in num­ber of en­tries, we feel that 2.5 months might be too short, so this round will run un­til end of this year.

We are look­ing for tech­ni­cal, philo­soph­i­cal and strate­gic ideas for AI al­ign­ment, posted pub­li­cly be­tween July 15 and De­cem­ber 31, 2018. You can sub­mit links to en­tries by leav­ing a com­ment be­low, or by email to ap­ply@ai-al­ign­ment.com. We will try to give feed­back on all early en­tries to al­low im­prove­ment. Another change from pre­vi­ous rounds is that we ask each par­ti­ci­pant to sub­mit only one en­try (though pos­si­bly in mul­ti­ple parts), rather than a list of sev­eral en­tries on differ­ent top­ics.

The min­i­mum prize pool will again be $10,000, with a min­i­mum first prize of $5,000.

Thank you!