A New Blackboard For The Poor, Via Rice University
"Can you ustnnderad this stnecnee?" Well, if you can, then you understand probabilistic logic -- and you're half-way to understanding the I-Slate being developed at Rice University by computing professor Krishna Palem.
The I-slate is a solar-powered LED tablet Palem and his project partners are hoping will replace the slate tablets and chalk currently used by school children in many poorer parts of the world. The I-slate uses a processing chip that requires as much as 30 times less electricity than a standard chip by using probabilistic logic (which finds estimated, rounded off answers) rather than Boolean logic (which finds precise answers).
For example, 7 + 3 = 10. In precise terms, 10 is the correct answer. But 7 + 3 = 11 is also a correct answer in probabilistic logic terms. It's a less correct answer than 10, but a more correct answer than 12. Kids will learn by searching for the most correct answer, rather than the only possible answer.
The slate is still in development, but the first school that will get to try it out has already been chosen. It's a school in rural India that has very few resources.
"There are a fair number of schools around the world where access to resources like reading and writing material, teachers, and electricity are not at all things you can take for granted" Palem tells Hair Balls. "In this particular school, for example, they do not have any access to electric power, and they have one [volunteer] teacher for what we would call grades one through five. You can imagine the learning challenges for the students."
Palem and the group at Rice University are just one of three groups working cooperatively on the project. Palem is providing the technical aspects, another group is working on getting the I-slates to the school in India, and another is working on a curriculum for "the box" (as Palem calls it.) That's an especially important part of the project because along with replacing the slate tablet and chalk, the I-slate also hopes to replace the teacher through specialized content. How's that work?
"I hate to use this word but people like to call it intelligence. I don't want to call it that, because it's misleading. I want to call it machine learning. There are systems that observe your behavior, like Amazon, and then they can make suggestions. If you like this movie, maybe you will like this movie, too. That technology is actually using a kind of a learning model. That model can be applied to a variety of situations.
"What we hope to accomplish is a level of literacy comparable to that of a decent school. Not a fancy one," he adds quickly, "just a decent school. One that has the right number of teachers, has electricity and books."
There are schools in the United States that also lack teachers and books. Why not give the I-Slate to some of them, instead of India?
"That's actually a very important question. I don't think we should avoid it. This is a test or a laboratory. The lab just happens to be in India. Once the concept is proved, it will become more widely available.
"We have what we can consider is a pyramid of applications. A pyramid, as you know, has a very broad base and gets more pointed at the top. At the bottom of the pyramid, it's helping a lot of people with very simple things.
"At the top, you have fewer people. They want a more fancy version of this, are not as constrained by cost. They want to save energy, but don't have to live off of solar energy. Of course, the expectation they will have is higher." (Like say, 7 + 3 = 10 and only 10, maybe?)
"Why not start with a school in Arkansas? I believe that needs are different. A child in Arkansas would need a slate more in the middle of the pyramid, not the bottom and not the top, somewhere in the middle. We are first focusing on the bottom of the pyramid, on the most demanding case.
"When the Wright Brothers flew the first plane, it happened at one place and there was one plane, but they were proving a point. We are also hoping to prove a point," Palem says. "Nothing would make us happier than if this was used widely, used, truly, everywhere. That's our ambition, our hope. Time will tell."
(Is that actual time or probabilistic time?)