Here is the paper. http://edge.org/3rd_culture/hillis04/hillis04_index.html
I agree with the general direction this paper points out. However, there are several important points I have to point out.
First of all, this paper was written in 2000 (very impressive to write something like this in 2000). Now in 2006, as a new wave of innovation surges in web2.0 , I don’t know if Danny Hillis will think differently now.
I did more than 3 years of research on elearning, working on a project very much like the concepts outlined in the core part of this paper about knowledge web. What we were trying to do is to design a learning object system to adapt to different learners according to his learning background, learning styles and so on. There are many similar attempts in both academic and industry fields. For now, let’s give them a name for convenience of this article: Instruction Design Based and AI Assisted Adaptive Intelligent Tutoring System (IAAITS).
I generally agreed that this approach holds one major hope that elearning has in changing dramatically the way people learn. Web will become a very important learning platform, which will making going to schools for so many years unnecessary. But as I was designing and implementing such kind of system, a lot of problems emerge and thus I had to dig deeper into these ideas. Some problems are technical such as the use of AI. But some are fundamental flaws of the ideas in IAAITS and thus need to be revised. (Here is a link to the abstract of my thesis: http://edt.missouri.edu/Fall2005/Thesis/LiuY-051706-T3789/short.pdf, If you want Introduction of my thesis or the complete one, let me know. My contact is in About Me page. My thesis realized those fundamental flaws associated with such kind of approach, and thus approach elearning with a different methodology)
What I consider fundamentally wrong with IAAITS :
1. for an active learner, at a certain time, what he wants to learn and how he wants to read the material is totally random. He might try many different ways at the same time. So it is pointless to have a system to render learning material according to some Instructional Design theories. Furthermore, it inhibits such learners to learn freely.
2. our goal is to promote active learning. But if we select all the material for the learner to learn, we are making the learner passive. Leaners need to learn how to learn. That is important. A big problem of schooling is that most people coming out of so many years of school still don’t know how to learn.
As most elearning system tried very hard to fantasize the way of teaching, elearning is really not about teaching. It is more about learning. It is about putting learning into the hands of learners. (To really understand this, you can read about unschooling and Mr. John Taylor Gatto’s site about the history of school:http://www.johntaylorgatto.com/)
In the way of fantasizing teaching, they certainly need to rely a lot on AI. But AI at the current stage is complex and unable to provide a good result at this point. Without AI, just by putting learning back to the learner, we are actually able to come up with a great system that will help learners tremendously (even much better than what AI promised but couldn’t achieve). The key of elearning is how it can help individual learners to learn by themselves. The new revised approach (for now, let’s call it Learner Centered Active Learning Platform, e.g. LCALP), relies on leveraging the web as a cooperation platform. It is in tune with web2.0 development.
AI hasn’t been providing us with great results yet. (Anyone can name some?) But if we apply AI on a different basis, e.g. on the whole web globally, AI can be put into great use. AI on this basis is very different from AI on merely processors or machines. AI on the whole web (especially web2.0) has receptors, which are those individuals who input their feelings into web. It has also actors and can get feedback. It collects consciousness and intelligence contributed by billions of individuals across the world. By putting so many consciousness and intelligence together, we will have a web brain. In essence, the whole planet will function as one: a biological creature with billions of individuals on this planet as its cells. In short, the life on this planet evolves. I wish a smart guy like Mr. Danny Hillis can work on this great thing. He seems to be on this mission for his life. 🙂 Web2.0 is a big step towards it. Web2.0 is essentially a mechanism for cells to fully connect. Also as Mr. Danny Hillis outlined, there will be rich content available online, which is vital for the success of web based learning platform. Web2.0 provides a very easy way for people to participate and contribute content. We will soon have very rich content in various formats. Semantic web will also be put into great use, probably in a different way than it is geared towards now.
Yes, knowledge web will soon become true, and lets embrace it. It is the beginning of a great era.
I like Danny Hillis’ work.Computer science combined with biology is of great interest to me. As I was learning biology/life science in my undergraduate and graduate study, I was fascinated by it. My most favoriate parts are brain and genetics. Studying biology is like getting direct touch with the great truth and thus there are a lot things we can learn from it.
The same as Danny Hillis, one big topic of my life is also how to combine creativity and business. As I think web2.0 will change the way we organize, we will certainly change the relation between creativity and business.
Oh, yeah, he is homeschooled.
(When it comes to elearning or learning, I always have too many things to tell, thus unable to tell. I guess by commenting on someone else’ papers, I can give my ideas some background, which will make them a little easier to understand. But a single article really cannot cover all the ideas about this new learning platform. I will have to write many small essays on it.)