No Picture
Featured Articles

List of multipolar research projects

This list currently consists of research projects suggested at the Multipolar AI workshop we held on January 26 2015. Relatively concrete projects are marked [concrete]. These are more likely to already include specific questions to answer and feasible methods to answer them

No Picture
AI Timelines

Discontinuous progress investigation

We are collecting cases of discontinuous technological progress, to inform our understanding of whether artificial intelligence research is likely to undergo such a discontinuity. This page details our investigation. Details Motivations We are interested in whether artificial intelligence research

No Picture
AI Timeline Surveys

AI Timeline Surveys

We know of thirteen surveys on the predicted timing of human-level AI. If we collapse a few slightly different meanings of ‘human-level AI’, then before adding the most recent survey (2016 ESPAI, results pending): Median estimates for when there will be

No Picture
Featured Articles

Cases of Discontinuous Technological Progress

We know of two technologies which underwent a discontinuous jump in improvement equivalent to more than several decades at previous rates. These are explosives and superconductors. We know of two more cases of smaller discontinuities, and several promising cases to investigate

No Picture
Featured Articles

Discontinuity from Nuclear Weapons

Nuclear weapons were a radical improvement over previous weapons in terms of energy released per weight of explosive (relative effectiveness). They do not appear to represent a substantial discontinuity in terms of cost-effectiveness of explosives, though the evidence there is weak.

No Picture
AI Timeline Surveys

Hanson AI Expert Survey

In a small informal survey running since 2012, AI researchers generally estimated that their subfields have moved less than ten percent of the way to human-level intelligence. Only one (in the slowest moving subfield) observed acceleration.

No Picture
Clarifying concepts

Human-Level AI

‘Human-level AI‘ refers to AI which can reproduce everything a human can do, approximately. Several variants of this concept are worth distinguishing. Details Variations in the meaning of ‘human-level AI’ Considerations in specifying ‘human-level AI’ more precisely: