Unprecedented Technological Risks

Report by Seán Ó hÉigeartaigh, Nick Bostrom, Nick Beckstead, William MacAskill, Neil Bowerman, Owen Cotton-Barratt, Toby Ord
Published on 01 September 2014

Executive Summary

The development of nuclear weapons was, at the time, an unprecedented technological risk. The destructive power of the first atomic bomb was one thousand times greater than other weapons; and hydrogen bombs increased that destructive power one thousand-fold again. Importantly, this technological development was extremely rapid. The bombing of Hiroshima and Nagasaki came a mere six years after Einstein’s initial warning letter to President Roosevelt. Nuclear technology created a significant risk of the deaths of hundreds of millions, which was openly acknowledged, with President John F. Kennedy later putting the odds of a nuclear holocaust at “somewhere between one out of three and even.”

In the near future, major technological developments will give rise to new unprecedented risks. In particular, like nuclear technology, developments in synthetic biology, geoengineering, distributed manufacturing and artificial intelligence create risks of catastrophe on a global scale. These new technologies will have very large benefits to humankind. But, without proper regulation, they risk the creation of new weapons of mass destruction, the start of a new arms race, or catastrophe through accidental misuse. Some experts have suggested that these technologies are even more worrying than nuclear weapons, because they are more difficult to control. Whereas nuclear weapons require the rare and controllable resources of uranium-235 or plutonium-239, once these new technologies are developed, they will be very difficult to regulate and easily accessible to small countries or even terrorist groups.

Moreover, these risks are currently underregulated, for a number of reasons. Protection against such risks is a global public good and thus undersupplied by the market. Implementation often requires cooperation among many governments, which adds political complexity. Due to the unprecedented nature of the risks, there is little or no previous experience from which to draw lessons and form policy. And the beneficiaries of preventative policy include people who have no sway over current political processes — our children and grandchildren.

Given the unpredictable nature of technologica progress, development of these technologies may be unexpectedly rapid. A political reaction to these technologies only when they are already on the brink of development may therefore be too late. We need to implement prudent and proactive policy measures in the near future, even if no such breakthroughs currently appear imminent.

View report

Subscribe to our mailing list to get our latest updates