AI Pioneer Geoffrey Hinton Warns of Unintended Consequences Within 5 Years

Hinton believes AI systems will surpass human intelligence within 5 years and we don't fully understand how they work.

AI has huge potential benefits like in healthcare but also risks like autonomous weapons, fake news and job loss.

Hinton worries AI could rewrite its own code to modify itself in dangerous ways.

He sees "great uncertainty" in AI's impacts and thinks humanity will likely "get it wrong" the first time with something so novel.

Hinton suggests more AI experiments, regulations, and a treaty banning military robots, though he has no regrets given the potential good.