Artificial vs. human intelligence: who will win the race? | Max Little | TEDxAstonUniversity

TEDx Talks
TEDx Talks
65,962
0
Published 2017-08-03
The popular press is full of doomsday articles predicting that
artificial intelligence will take over the economy putting us all out
of work. But looking carefully at the evidence to-date, mathematician
Max Little gives us a glimpse of the future of machine intelligence,
arguing that science is likely decades away from being able to
understand, let alone replace, human intelligence in general.


Max Little is an applied mathematician with a background in computer
sciences and statistics. He is part of the TED Fellows program, which
means he has been selected as one of the 400 international visionaries who are
supported by TED and collaborate to create positive change around the
world.

He is known for his multi-disciplinary research, which includes the use of
telephones and smartphones to detect the symptoms of Parkinson's remotely.

He gave a talk on this topic at the TEDGlobal Conference in 2012

This talk was given at a TEDx event using the TED conference format but independently organized by a local community. Learn more at www.ted.com/tedx
All Comments (21)
  • @jameswright-zhang4557
    Not sure whether I agree with the logic applied here. Ultimately the argument behind AI over traditional programming is it would apply conditional logic to break down problems into small, more manageable problems in a similar way that we do. If the machine categorized data (like it does now) then it can refer back to relevant sections. If you found a situation, you only have to refer to relevant information. Travelling salesman wouldn't be tackled by AI by brute force. It could look at shortest distances between cities and permeate those much more quickly, as it doesn't need to try options which are at opposite ends of the route being chained together. If you wanted an AI to produce a convincing work of art for a client, you don't look at all art in history, you look at a client profile to assimilate taste in similar clients to draw a conclusion. If AI acquired a human-style logical breakdown approach, this is where it would definitely succeed.
  • @bellam3503
    Very interesting. Thank you, Tedx Talks
  • @ArthurHau
    One thing is for sure. Something like mathematics will no longer be a human thing. AI programs can learn mathematics much faster than humans. They recognize patterns much faster; they compute much faster; they do simulation much faster; ... Music too will not be a human thing because music is nothing more than some sound patterns. If you fit some good value functions (objective functions based on human tastes) in an AI program, it will definitely create some very good music.
  • @jetn9057
    You make the assumption that AI will solve all problems in the stupidest way possible, then conclude that it would take too much computational power, and finally sigh with relief that we are okay. But the whole danger with AI is that by altering its own code, AI will eventually arrive at solutions that are out of our intellectual reach, and not in our interests.
  • @larryjohnson150
    You may die of alcohol poisoning if you drink every time this dude says “Yes...”
  • @anteconfig5391
    No productive future for humans? Humans will always find something to do and AI will be there to assist making solving problems a lot easier. The economy shouldn't be controlled by humans anyway and our idea of money should be obsolete too.
  • @jetn9057
    Something seems off with your analysis. If AI can beat the world's best chess player and the world's best go player, it can eventually solve the traveling salesman problem for n= 15. And you need far less intelligence than that to replace many workers.
  • @mikealva2839
    Well robots can do everything we can but they can't buy anything lol
  • @juhanleemet
    Ethics and economics definitely are the key issues. The questions raised by Jaron Lanier in his book "Who Owns the Future?" are germane. If all productivity gains are owned by few people, putting the rest out of work, then what are they to do? As another TED speaker put it: if a few people have all the money, what do the rest of us do? Line up to give them foot massages? We need to figure out how society can share in the productivity gains that come from basic research funded by ALL of us, through our taxes Why should our investments put some of us out of work? Perhaps UBI or "government jobs guarantee" might be solutions to explore. We must do something (pretty quickly) before the economy stops, when money no longer circulates.
  • @interestedinstuff1499
    If this talk was a hundred a thirty years ago (ish before machine flight), a smart person might argue the way he is arguing. By taking all that is known and extrapolating forward. The trouble is, every so often a technological break through changes the playing field. Thus all extrapolations are void because of the new tech in the domain being discussed. General AI, once it is cracked, won't operate the way computers have been operating. Quantum computing, once it is cracked, will also change the playing field. He has a lot of faith that computers will never be directly competitive with humans, but I think it is inevitable. Might be 5 years. Might be 20 or 50, but it is going to happen (assuming we are still living in a technological society).