It’ll never work! A collection of experts being wrong about what is technologically feasible.
GiveWell update on its investigations into global catastrophic risks. My biggest disagreement is that I think nano-risk deserves more attention, if someone competent can be found to analyze the risks in more detail. GiveWell’s prioritization of biosecurity makes complete sense given their criteria.
Online calibration test with database of 150,000+ questions.
An ambitious Fermi estimate exercise: Estimating the energy cost of artificial evolution.
Nautilus publishes an excellent and wide-ranging interview with Scott Aaronson.
Video: robot autonomously folds pile of 5 previously unseen towels.
Somehow I had previously missed the Dietterich-Horvitz letter on Benefits and Risks of AI.
Robin Hanson reviews Martin Ford’s new book on tech unemployment.
Heh. That “stop the robots” campaign at SXSW was a marketing stunt for a dating app.
Winfield, Towards an Ethical Robot. They actually bothered to build simple consequentialist robots that obey a kind-of Asimovian rule.