Rating is available when the video has been rented.
This feature is not available right now. Please try again later.
Published on Jul 30, 2011
James Newton-Thomas speaks about AI, robotics, super intelligent AI, the Singularity, and friendly AI
Quote:"The biggest change in AI that has changed in the last 20 years, is that people have given up trying to define the problems that they wish to solve. They have realized that the problems could often be so complex that we may not know a solution, yet we can still develop systems, or evolved systems that can function in a complex environment."
"Neural Nets have, endemic to their operation, a type of evolutionary paradigm. But you can then go to a meta-paradigm, which is to change the structure of the network, and to change the types of neurons and the interconnections between them in the same way that nature will specialize - our brains contain a myriad of sub-brains, each one architecturally adept at solving a certain class of problem."
One of the interesting aspects of the interview is how AI has changed, problems can be so complex that it is so hard to define a scalable way to deal with the problem - but we have realized that the problem does not have to be fully described. By developing systems that operate in an evolutionary paradigm, it is possible to solve classes of problems that we don't know how to define.