At last week’s Aspen Security Forum, former Google CEO Eric Schmidt compared artificial intelligence to nuclear weapons. He spoke about the dangers of AI and called for technology to be more in line with the ethics and morality of the people it serves, reports Vice.

In Schmidt’s opinion, in the near future the US and China should conclude an agreement on the use of AI, as they did in the 50s and 60s of the last century for nuclear deterrence. However, it took a decade of testing and the destruction of two Japanese cities, Hiroshima and Nagasaki, before humanity passed a series of laws that restrained the use of nuclear weapons.

Then nuclear weapons killed tens of thousands of people and showed the world their terrible potential. Humanity does not currently use it because of mutual assured destruction – the theory of deterrence, which states that if one country launches nuclear weapons, another will do the same, and civilization may perish.

Although Schmidt compares AI to nuclear weapons, humanity hardly needs “mutually assured destruction” to use it. AI does not have the destructive power of nuclear weapons. It also reflects the values ​​of its creators and is only as good as the people who developed it. Therefore, for example, chatbot trained on 4chan’s toxic image board becomes a hate speech machine.

This is the opinion of Demis Hassabis, CEO of DeepMind, which trained artificial intelligence to defeat players in Starcraft II. 

“It matters who builds [AI], which cultures they come from and what values they have, the builders of AI systems. The AI systems will learn for themselves…but there’ll be a residue in the system of the culture and values of the creators of that system,” says Hassabis.