I like it.

It’s funny but also informative.

The problem with using the ‘split-and-recombine’ technique is that people are mobilized from their idle position to cross the divide between black and white and there’s a ‘wearing out’ effect on people as they take these polarized positions and then have to come back to the middle. The ‘wearing out’ effect means that you can do it only so many times before you lose your audience. But isn’t there an infinite supply of people to listen to polarized messages?

Basically, I believe there is a way to achieve the same effect (i.e. generate a hype wave) without having to do what I did.

It’s like the difference between Flash Memory and MRAM. In Flash electrons move from ne place to another and this wears out the junction (or whatever structure) so you can only store things on a Flash card 100,000 times. We’re talking elementary particles here, not human beings. With humans beings I can’t think of anyone who can take this wearing out effect more than 20, 50 times before they drop out.

MRAM will eventually replace all RAM (when they can scale densities and production) and which will mean that your computer can be turned off (to save POWER/save us from oild dependence etc) and when you turn it on again everything, all your programs, will be just the way you left them. MRAM doesn’t wear out because the electrons don’t cross the junction back and forth, which is what happens in Flash. Thus, MRAM is more useful than Flash.

So I want to study the mechanism in MRAM ata high level and see if any wisdom can be extracted from its design.

Re: AI

By “AI” I meant simply “Inference Engines” which exist today and which already have been used with domain specific ontologies to reason about domain specific information.

I did not mean Mr. Data or HAL.

Also, with respect to the Global Brain, the definition there is again not HAL. It’s the complex behavior that emerges from the interaction of all P2P Semantic Web Inference Engines. The complex behavior is not al algorithm. It’s the result of massively parallel interactions between nodes (or Engines) with relatively simple behavior. It doesn’t imply intelligence will emerge. I did not clarify that in the Wikipedia article and that got some people carried away … 🙂 May be I didn’t clarify it intentionally but I tend to believe that I just didn’t realize that people would assume I was talking about an AI that emerges out of the complex behavior. I might have wanted the vaguness so that people start thinking (there is a lot to learn from the New Kind of Science book by Wolfram if people wanted to stretch their imagination but I definitely did not mean AI.. I mean complex behavior that can be thought of as a Global Brain in metaphorical way)

Yet, Inference Engines are considered AI Engines (expert systems use inference engines) and that is the kind and level of AI I meant.

…

Thank you for doing a well rounded analysis. I know how hard it is to wrap our minds around the big ideas, especially the “to split or not to split” debate ‘)

Marc