2016 – AI Impacts

2016 Expert Survey on Progress in AI

Published June 2016; last substantial update before Oct 2017 The 2016 Expert Survey on Progress in AI is a survey of machine learning researchers that Katja Grace and John Salvatier of AI Impacts ran in

Concrete AI tasks for forecasting

This page contains a list of relatively well specified AI tasks designed for forecasting. Currently all entries were used in the 2016 Expert Survey on Progress in AI. List Translate a text written in a newly discovered language

Joscha Bach on remaining steps to human-level AI

By Katja Grace, 29 November 2016 Last year John and I had an interesting discussion with Joscha Bach about what ingredients of human-level artificial intelligence we seem to be missing, and how to improve AI forecasts more generally. Thanks

Tom Griffiths on Cognitive Science and AI

This is a guest post by Finan Adamson, 8 September 2016 Prof. Tom Griffiths is the director of the Computational Cognitive Science Lab and the Institute of Cognitive and Brain Sciences at UC Berkeley. He

Conversation with Tom Griffiths

Participants Professor Tom Griffiths, Director of the Computational Cognitive Science Lab and the Institute of Cognitive and Brain Sciences at the University of California, Berkeley. Finan Adamson, AI Impacts. Note: These notes were

What if you turned the world’s hardware into AI minds?

By Katja Grace, 4 September 2016 In a classic ‘AI takes over the world’ scenario, one of the first things an emerging superintelligence wants to do is steal most of the world’s computing hardware and repurpose it to running the AI’s

Sources of advantage for digital agents over biological agents

Artificial agents should have several advantages over humans. Details The following is an excerpt from Superintelligence (Bostrom, 2014), reproduced with permission. It outlines ten advantages Bostrom expects digital intelligences to have over human intelligences. Sources of advantage for digital intelligence Minor

Examples of early action on risks

Details Discussion There are many current efforts to mitigate risks from artificial intelligence. We might learn something about the likelihood of these efforts influencing AI risk by looking at similar past efforts. To this end, we are interested here

Friendly AI as a global public good

By Katja Grace, 8 August 2016 A public good, in the economic sense, can be (roughly) characterized as a desirable good that is likely to be undersupplied, or not supplied at all, by private companies.

Costs of extinction risk mitigation

We very roughly estimate that the annual cost of reducing the probability of human extinction by 0.01% is within the range of $1.1 billion to $3.5 trillion. Introduction This article is intended to be usable