Simulations Poised to Accelerate Learning

March 12 2008 / by Venessa Posavec
Category: Metaverse   Year: 2008   Rating: 7

It was recently announced that hospitals in Isreal have begun using virtual reality programs to diagnose and treat patients with brain injuries. The patient tries to catch a virtual tennis ball being thrown on a screen, their actual hand movements are recorded, and the information is fed into a computer program. The program then diagnoses whether the person has had a traumatic brain injury or stroke (with 90-98% accuracy!), and run a series of simulations to determine what will be the most effective treatment and rehabilitation methods.

This is a huge step in demonstrating the value of virtual environments and simulations to do real good in the world. For almost all of the decisions we make, we run simulations in our brain without even thinking about it. “If I do A, then B is a likely and desired outcome.” Through trial and error, our simulations get more accurate over time – we may call it “wisdom”. But, in some situations, such as the above brain injury example, even our best human guesses for the right course of action may be wrong. By running computer simulations, we can take that guesswork out. Instead of creating a rehabilition therapy that may not only be ineffective, but downright harmful, doctors will now be able to implement the most effective therapies according to the patient’s level of injury.

As computer processing speed continues to increase, and we methodically quantify the underlying systems that drive everything around us, we’ll see simulations popping up as tools for increasing efficiency in all fields. I can see this being used to improve learning and skill development in both education and the workplace.

We just spoke with Katie Salen of the Parsons School of Design about a school she’s designing that’s centered around game-based education, using video games and virtual worlds. Imagine what will happen when we’re able to quantify the type of learning that happens during game play. Each student could have a customized educational program tailored to suit the way their individual brains processes information. Students will leapfrog through a K-12 education within a few years, and eventually, maybe even months.

In the workplace, the procedures manual and often cumbersome training program will be a thing of the past. New employees will interact with a virtual program that might initially test their decision-making abilities and skill sets. It might put that information through a program to quantify your managerial skills and deficiencies, run simulations of the best way to train you, and quickly mold you into a superstar.

By modeling the way the human brain works, a powerful new method of making decisions and increasing efficiency is forming. The difference is that the computer can run thousands more simulations that we could, do it in a fraction of the time, and base it on a wealth of information far greater than what is stored in one individual brain. It’s our nature to increase productivity while minimizing effort (at least economics would indicate that), and it seems we’re on the verge of doing this at a level never seen before in human history.

What other ways do you think virtual environments and simulations will be used to make the world a better place?

Comment Thread (2 Responses)

  1. Imagine what will happen when we’re able to quantify the type of learning that happens during game play.

    That’s a powerful idea, and not too far off I think. First, the technology needs to be made fool-proof and more affordable. Second, a large body of brain/learning/simulation data must be collected—which could happen en mass as kids play games wearing neural pattern recorders (perhaps participating in such Learning Logging would earn them some $ or at least subsidize their game play – I could see a large business model being built around this). Third, we’d need to make sense of this bank of patterns + all the correlations, then devise efficient tests that capture the more relevant brain data. Fourth, repeat steps 2 & 3 until actionable information / suggestions begins to emerge at a high/desired rate.

    Seems like this will require a great deal of computational power. Fortunately, as you point out, that’s on the way.

    Posted by: Alvis Brigis   March 12, 2008
    Vote for this comment - Recommend

  2. What I found most interesting in this release was the fact that a program used for the rehab of patients created a plethora of data from which a computer was able to “learn” how to diagnose a variety of types of brain trauma – an example of the complex AI/computational abilities that can spawn from just a bank of data and a controlled environment. It begs the question – what kind of complex AI will evolve simply from crunching the data we’re beginning to quantify?

    Posted by: Marisa Vitols   March 12, 2008
    Vote for this comment - Recommend