Mini-conference "Near-term AI safety"

TL;DR: The event will be in Moscow, Russia, and near-term risks of AI will be discussed. The main language will be Russian, but Jonatan Yan will speak in English from HK. English presentations will be uploaded later on the FB page of the group “Near-term AI safety.” Speakers: S. Shegurin, A. Turchin, Jonathan Yan. The event’s FB page is here.

In the last five years, artificial intelligence has developed at a much faster pace in connection with the success of neural network technologies. If we extrapolate these trends, AI near-human level may appear in the next five to ten years, and there is a significant probability that this will lead to a global catastrophe. At a one-day conference at the Kocherga rationalist club, we’ll look at how recent advances in the field of neural networks are changing our estimates of the timing of the creation of AGI, and what global catastrophes are possible in connection with the emergence of an increasingly strong AI. A special guest of the program Jonathan Yan is Hong Kong will tell (in English, via Skype) the latest research data on this topic.

The language of the conference: the first two reports in Russian, and the report Yan in English without translation, the discussion after it in English.

Registration: on the event page on Facebook.

Place: rationalist club “Kocherga”, mail hall, Bolshaya Dorogomilovskaya ul., 5 корпус 2.

Participation in tariffs anticafe, 2.5 rubles a minute, coffee for free.

Videobroadcast will be on the Facebook.

Program:

October 14, Saturday 15.00 - the beginning.

15.00 - Shegurin Sergey. “Is it possible to create a human level AI in the next 10 years?”

16.00 - Turchin Alexey. “The next 10 years: the global risks of AI before the creation of the superintelligence”

17.00 - Jonathan Yan. “Recent Developments Towards AGI & Why It’s Nearer Than You Think (in English)”

17.40 - Discussion

Mini-conference “Near-term AI safety”