Computers That Can Learn From Experience

Interesting article in the New York Times.

Computers have entered the age when they are able to learn from their own mistakes, a development that is about to turn the digital world on its head.

[…]

In coming years, the approach will make possible a new generation of artificial intelligence systems that will perform some functions that humans do with ease: see, speak, listen, navigate, manipulate and control. That can hold enormous consequences for tasks like facial and speech recognition, navigation and planning, which are still in elementary stages and rely heavily on human programming.

So let me get that straight. Most humans can’t learn from their own mistakes and yet we can get computers to do it? 😉

Seriously though, this is going to affect education in possibly two ways.

First, it’s possible that we will be able to use computers to test out new learning pedagogy. Imagine a teacher being able to have a virtual class of 30 simulated students. You can try out activities or tests on them and perhaps see how effective it they are.

Second, computers like this will be good teachers because they can adapt to changing circumstances. Not only is the software coming along rapidly, but the hardware required to do it has come down in cost and size. IBM’s Watson supercomputer (it won Jeopardy) used to take up a whole room but now it’s the size of three pizza boxes. IBM is so impressed with the technology they have spun off a new business unit to exploit it.

H/T Forbes

The company has created a stand-alone group to develop big data solutions, funded with $1 billion and staffed with 2,000 employees, and moved the entire shebang to a hip location in Manhattan’s East Village. Its very existence will change conversations about big data, even those with customers in which Watson doesn’t participate, and it will be hard particularly for big clients to believe promises of the magic of the crowd when Big Blue is willing to map out (and deliver) the work behind it. Watson could also take its demonstrated prowess at winning Jeopardy! and make some wildly novel or meaningful insights into newsworthy, topical issues, thereby demonstrating both the promise and peril of big data reality, not its buzz.

Watson exists to eat big data and to generate insights from it. What if it got data about a student? I have speculated we will soon see Smart Toys which will always watch their children and generate data on them. Perhaps the technology will soon exist to help students learn independently from teachers? This seems like a big step in that direction.

 

Btw, I am still going to make predictions for 2014 but I’m having problems this year with my scenerios. I’ll see if I can put it up on Friday.

Advertisement