Will machines outsmart us on all tasks within decades, as many industry leaders believe? Max Tegmark, MIT physicist and researcher, president of the Future of Life Institute and the author of the new NY Times Bestseller “Life 3.0: Being Human in the Age of Artificial Intelligence” presented the most provocative ideas from his book to the Web Summit audience.
Discussing our collective journey into the future, Tegmark compared the development of a rocket to AI, with a focus on how powerful it is, how it can be steered and where we want it to go.
During the past 13.8 billion years, the universe has transformed from “dead and boring” to progressively more complex, right up to organisms like ourselves with a passion for life. Tegmark said this was humbling, as, compared to the scale of the cosmos, life is tiny, but also inspiring because now, thanks to technology, it doesn’t have to remain tiny.
Tegmark said that, after basic life, humans represented Life 2.0 because they can learn, with the future technological merge acting as Life 3.0. He discussed his hatred of “carbon chauvinism”, by which people believe only carbon based lifeforms can be clever.
He demonstrated how an algorithm playing an Atari game learns a trick and shamelessly exploits it in a way that would be very impressive if learned by a four year old in the same amount of time.
These days most researchers accept that machines will be able to do everything that humans can, which some attribute as scaremongering, but Tegmark calls “safety engineering” as he believes it is important to consider all things that could potentially go wrong to make sure everything goes right.
In his opinion, we can win the wisdom race by:
- Banning lethal autonomous weapons (the machines which automatically decide who to kill with no human input)
- Ensuring AI generated wealth makes everyone better off and finds new ways of helping people, as opposed to harming them
- Investing in AI safety research
- Thinking about what sort of future we want
If AI amplifies intelligence, Tegmark said it might allow humanity to flourish like never before, but in order to do this, we need to figure out how to teach machines the collective goals of humanity. The question is how to live with super intelligence and what sort of future we want to create.