AI Impacts

AI Timeline Surveys

This page is out-of-date. Visit the updated version of this page on our wiki. Published 10 January 2015 We know of twelve surveys on the predicted timing of human-level AI. If we collapse a few slightly different meanings of

Michie Survey

In a 1972 poll of sixty-seven AI and computer science experts, respondents were roughly divided between expecting human-level intelligence in 20 years, in 50 years and in more than 50 years. They were also roughly divided between considering a ‘takeover’

AI and the Big Nuclear Discontinuity

By Katja Grace, 9 January 2015 As we’ve discussed before, the advent of nuclear weapons was a striking technological discontinuity in the effectiveness of explosives. In 1940, no one had ever made an explosive twice as

The Biggest Technological Leaps

By Katja Grace, 9 January 2015 Over thousands of years, humans became better at producing explosions. A weight of explosive that would have blown up a tree stump in the year 800 could have blown

The AI Impacts Blog

By Katja Grace, 9 January 2015 Welcome to the AI Impacts blog. AI Impacts is premised on two ideas (at least!): The details of the arrival of human-level artificial intelligence matterSeven years to prepare is very different from

Cases of Discontinuous Technological Progress

We know of ten events which produced a robust discontinuity in progress equivalent to more than one hundred years at previous rates in some interesting metric. We know of 53 other events which produced smaller

Effect of nuclear weapons on historic trends in explosives

Nuclear weapons constituted a ~7 thousand year discontinuity in relative effectiveness factor (TNT equivalent per kg of explosive). Nuclear weapons do not appear to have clearly represented progress in the cost-effectiveness of explosives, though the

AGI-09 Survey

Baum et al. surveyed 21 attendees of the AGI-09 conference, on AGI timelines with and without extra funding. They also asked about other details of AGI development such as social impacts, and promising approaches. Their findings include the following:

Bainbridge Survey

A survey of twenty-six technology experts in 2005 produced a median of 2085 as the year in which artificial intelligence would be able to functionally replicate a human brain (p344). They rated this application 5.6/10 in beneficialness to

AI@50 Survey

A seemingly informal seven-question poll was taken of participants at the AI@50 conference in 2006. 41% of respondents said it would take more than 50 years for AI to simulate every aspect of human intelligence, and 41%