In this short talk, Future of Humanity Institute researcher Stuart Armstrong has 5 minutes to lay out the reasons why Artificial Intelligence, if possible, will likely be extremely powerful and considerably dangerous (though not at all in the ways movies imply). The talk was given at the 25th Oxford Geek Night.

A global catastrophic risk is one with the potential to wreak death and destruction on a global scale. In human history, wars and plagues have done so on more than one occasion, and misguided ideologies and totalitarian regimes have darkened an entire era or a region. Advances in technology are adding dangers of a new kind. It could happen again.

Oxford philosopher and transhumanist Nick Bostrom examines the future of humankind and asks whether we might alter the fundamental nature of humanity to solve our most intrinsic problems.

Eliezer Yudkowsky, Research Fellow for the Singularity Institute, on cognitive biases and giant risks.