Saturday September 16, 2017

Recipes for survival in a world with superintelligent artificial intelligence

(Credit: iStock/Getty Images)

Listen 13:49

Who's worried about Superintelligent AI?
Physicist Max Tegmark is part of a group of intellectuals and futurists, including luminaries like Stephen Hawking, Elon Musk and Bill Gates, who have sounded the alarm about the existential risks to humanity of superintelligent artificial intelligence. Dr. Tegmark, a physicist at MIT and president of the Future of Life Institute, thinks the public is largely worried about the wrong things, and we need to set an agenda to study how we can safely live with AI to avoid doomsday scenarios, and reap the great potential benefits it could bring.

There's no law of physics that says one can't build machines more intelligence than us, in all ways.  So I really don't like this sort of carbon chauvinism that says AI can't be as smart as us, and I think we've only seen the tip of the intelligence iceberg.  There's really this amazing potential to unlock the full intelligence that's latent in nature and use it to help humanity flourish or screw up in new ways. - Dr. Max Tegmark

The Book

Life 3.0 cover
Dr. Tegmark's new book is Life 3.0 - Being Human in the Age of Artificial Intelligence. In it he explores why he's confident that, sooner or later, we will develop an artificial intelligence that will be far smarter than we are - and that's just the beginning. AI will revolutionize our economy, drive technological innovation and generally transform our world. The big question he has is whether we're smart enough to make sure AI is created so that the machines' goals are compatible with our goals. He thinks we're distracted by Hollywood's fantasies of intelligent killer robots, when we need to be concerned with practical problems of how to design - and live with - AI in a way that serves our interests.  

We really need to for starters to figure out how we can transform todays buggy and hackable computer systems into robust AI systems that we really trust.  These are research questions and we should focus a lot of research effort on them so we get the answers by the time we need them. - Dr. Max Tegmark

The Future

neural net illustration

Neural net illustration

Dr. Tegmark thinks the next 100 years will be the tricky part to both predict and successfully navigate. Beyond that he's pretty sure that AI will be the way we enter the future - and even expand our culture - or whatever our culture evolves into.  And ultimately he predicts very exciting things for the next billion years or so.

There is just an incredible potential for the future of life, and most science fiction writers have been much too pessimistic in what we can ultimately do, because they've been so limited to this idea that if you want to go far away you need to send a bunch of meat bags there ... In fact, ultimately life is information and the much smarter way to do it is to basically email the information about your mind ... and have yourself reassembled there.  - Dr. Max Tegmark