“AI is going to be both extremely unpredictable and unimaginable… what do you do about that the problem with AI is that it is so impactful, it is so powerful, it can solve everything, but it can also do everything, and all these questions don’t have answers right now.” — Ilya Sutskever
i was born in Russia and my parents immigrated to Israel when I was five i was going to school and I was a good student and my parents were looking for an environment where I could learn more and at some point by sheer chance we stumbled upon the Open University and I started taking classes in the 8th grade it was the best experience possible i got my books by the way the Open University books are very clearly written i just got those books and I read them and I understood to the point where I remember there being a before and an after i became confident that if I just read something very slowly I would eventually understand it so that was very helpful also you know I studied math and computer science i got a very I got a very strong foundation when my parents moved to Toronto one of the first things I did was to go to the Toronto Public Library and try to find a book on machine learning rather than finishing high school once again what I tried to do was to become a transfer student to the University of Toronto the great stroke of luck of being in Toronto is that Jeff Hinton was there this was the place to be those were the most forward-looking ideas in AI in the world back then and so I was able to join the University of Toronto as a transfer student in 2002 computers could play computer games a little bit and they could play chess and checkers and that was it and I remember thinking a very big computer could play chess no problem but how can it learn how is learning possible at all can computers learn like somehow I felt that if you had the answer to the learning question then everything else would follow as a result of the Alexent paper that we wrote in grad school some companies expressed in a desire to acquire our company which didn’t exist so we needed to create a company and eventually Google acquired us and so I joined Google the idea that the large neural network can do anything was further supported by some of the research we’ve done at Google and then the opportunity to start open AI has presented itself i was in the Bay Area it felt like how can I be here and not really try a real like a real serious startup with all these illustrious people that came together so I decided to go for it we continued working at OpenAI deep learning is a big idea and quite a few people contributed to it and I’m happy that I was able to contribute to it as well but with AI you know unfortunately the future is not so simple ai is exciting because it’s powerful right you have the power to do stuff what kind of stuff would you imagine the AI would do if the AI became powerful enough if the AI became capable enough we’ll have incredible health care what if an AI was doing medical research that would be amazing you could do so much more you could cure so many diseases and maybe extend life i think those are really wonderful things but if an AI can do this what else can it do and ultimately the answer is going to be everything one of the challenges is that AI is going to be both extremely unpredictable and unimaginable it’s unimaginable how can we prepare we need to prepare it’s not clear how and then eventually the power of AI is going to be so vast the idea that an AI can build the next generation of AI the intelligence explosion holy moly what do you do about that the problem with AI is that it is so impactful it is so powerful it can solve everything but it can also do everything and all these questions don’t have answers right now i want to express my deepest gratitude to the open university for giving me this honorary degree i find it so meaningful because for quite some time open university for me represented all of academia and all of interesting learning it’s really closing a circle in in a way that’s very meaningful