Modeling AI milestones to adjust AGI arrival estimates?

post by Ozyrus · 2019-07-11T08:17:55.914Z · score: 11 (5 votes) · LW · GW · 2 comments

This is a question post.

Contents

  Answers
    4 Michaël Trazzi
None
No comments

Is there a roadmap of major milestones AI researchers have to achieve through before AGI is here? I feel like this would be beneficial to update on AGI arrival timelines and adjust actions accordingly. Lately, a lot of things that can be classified as a milestone have been achieved -- GLUE benchmark, StarCraft II, DOTA 2 etc. Should we adjust our estimates on AGI arrival based on those or are they moving according to plan? It would be cool to have a place with all forecasts on AGI in one place.

Answers

answer by Michaël Trazzi · 2019-07-11T11:53:55.952Z · score: 4 (3 votes) · LW · GW

You can find AGI predictions, including Starcraft forecasts, in "When Will AI Exceed Human Performance? Evidence from AI Experts". Projects for having "all forecasts on AGI in one place" include ai.metaculus.com & foretold.io.

comment by moridinamael · 2019-07-12T04:43:26.973Z · score: 3 (2 votes) · LW · GW

Looks like all of the "games"-oriented predictions that were supposed to happen in the first 25 years have already happened within 3.

edit: Misread the charts. It's more like the predictions within the first ~10 years have already been accomplished, plus or minus a few.

2 comments

Comments sorted by top scores.