Blog

Jan 22, 2016

How ‘artificial swarm intelligence’ uses people to make better predictions than experts

Posted by in categories: information science, robotics/AI

I have seen this model so many times over the decades. And, I even was engaged in some of these experiments in the past. The continued problem we saw is “subjective reasoning” by humans which makes the experiments flawed.

And, as Yampolskiy suggested this is not true AI; it is using human insights and identifying patterns based on human input whch also includes subjective reasoning.


While AI focuses on creating intelligent machines that perform human tasks, a human-based algorithm, harnessing the power of the crowd to make predictions, shows remarkable accuracy.

Read more

4

Comments — comments are now closed.

  • Victoria Jones on January 25, 2016 1:37 pm

    Hi Karen,
    It’s true that traditional Wisdom of Crowds has not evolved much over the decades, but this is different. It’s not Wisdom of Crowds (which aggregates individuals in isolation), it’s Swarm Intelligence (which builds synchronous closed-loop systems to foster a higher-level emergent intelligence).

    Yes, the building blocks are sentient and capable of subjective reasoning but that’s a good thing. Research shows that swarms can exceed natural human abilities while keeping human morals and values in the loop. From this perspective, it could be the safest path for exceeding natural human intellect.

    This interview by Singularity 1-on-1 gives a good introduction to the concept: https://www.youtube.com/watch?v=OawkMleAEdQ

    -Victoria Jones

  • Barnabas on January 27, 2016 5:31 am

    I agree with @Victoria, swarming is very different than the prior methods for dealing with wisdom of crowds.

    Prior methods collect and crunch data, generating a snapshot of group opinions. Swarming, on the other hand, creates a living system that enables emergence.

    This is why swarming is A.I., and could be the safest path to super-human sentience.

  • Karen on January 27, 2016 6:50 am

    I just do not agree

  • Karen on January 27, 2016 7:07 am

    It is a flawed technique. Seen it too many times. Therefore, nice try, I do not agree. End of story