Do the risks of human extinction due to the coming growth of Artificial General Intelligence (AGI) combined with the Internet of Things (IOT) rank at the top of the list of existential risks we now face, along with the worst risks from climate change and the risks which still exist from misuse of nuclear technology and biotech?
On World Futures Day, Vint Cerf (father of the internet) gave an excellent overview of current policies on internet risks, which generally assume that we are far away from the time when AGI might be as dangerous as some fear. Paul Werbos, who was recently awarded the IEEE Frank Rosenblatt Prize for two of the underlying technologies behind the recent revolution in deep learning, and more powerful AI, said that extreme changes in technology -- both risks and opportunities -- are far greater than most decision makers and policy specialists in this area begin to know.
The diversity of views is based a lot in the sheer fragmentation and proliferation of work in this field, due to the many immediate applications of simpler internet and AI technologies. He urged us to look closely at
http://1dddas.org/activities/
to learn just how great the diversity of views and knowledge really is. What one expert group assures us is far away has already been implemented in working physical engineering systems in another. In 2014,
the most advanced conference in the field, the IEEE World Congress on Computational Intelligence, invited Werbos to speak in several plenary sessions and submit a paper, https://arxiv.org/abs/
https://www.sciencedirect.com/
No comments:
Post a Comment