Blog

GoCAS talk on AI Impacts findings

By Katja Grace, 27 November 2017 Here is a video summary of some highlights from AI Impacts research over the past years, from the GoCAS Existential Risk workshop in Göteborg in September. Thanks to the folks there

Blog

Price performance Moore’s Law seems slow

By Katja Grace, 26 November 2017 When people make predictions about AI, they often assume that computing hardware will carry on getting cheaper for the foreseeable future, at about the same rate that it usually

AI Timelines

2017 trend in the cost of computing

The cheapest hardware prices (for single precision FLOPS/$) appear to be falling by around an order of magnitude every 10-16 years. This rate is slower than the trend of FLOPS/$ observed over the past quarter century,

No Picture
Hardware and AI Timelines

Price-performance trend in top supercomputers

A top supercomputer can perform a GFLOP for around $3, in 2017. The price of performance in top supercomputers continues to fall, as of 2016. Details TOP500.org maintains a list of top supercomputers and their

No Picture
AI Timelines

Computing hardware performance data collections

This is a list of public datasets that we know of containing either measured or theoretical performance numbers for computer processors. List Top 500 maintains a list of the top 500 supercomputers, updated every six

No Picture
AI Timeline Surveys

2016 ESPAI Narrow AI task forecast timeline

This is an interactive timeline we made, illustrating the median dates when respondents said they expected a 10%, 50% and 90% chance of different tasks being automatable, in the 2016 Expert Survey on progress in

No Picture
AI Timeline Surveys

Automation of music production

Most machine learning researchers expect machines will be able to create top quality music by 2036. Contents DetailsEvidence from survey dataSummary resultsDistributions of answers to Taylor question Details Evidence from survey data In the 2016

Arguments for AI risk

Stuart Russell’s description of AI risk

Stuart Russell has argued that advanced AI poses a risk, because it will have the ability to make high quality decisions, yet may not share human values perfectly. Details Stuart Russell describes a risk from