Arimaa - The game of real intelligence

Vaguely interesting. Arimaa is a game, playable with a standard chess set, which is explicitly designed to be difficult for computers. There is a moderate prize ($10k), and some interesting theory behind it, but other than that, it seems easier to just use Go as a 'harder and larger search space' for computers.

Google's Desktop Search adds Sidebar

Most people know about this, yet another Google related way of adding functionality to your desktop. An update to Desktop Search adds Desktop Widgets, rather like Konfabulor or Mac OSX.

It's interesting, but it's at most a marginal update to the way I do most things. I move some functionality from Mozilla to the Sidebar(which is convenient, since Firefox is often on the edge of crashing with all I load it in) and get slightly better indexing. Good things, but not amazing.

For someone who hasn't yet gotten desktop search, I'd recommend it highly. Very clever, and less resource intensive than other solutions which is good for older computers.

Review by Ars Technica at the above link.

Inference Group: Dasher Project: Home

Fascinating. I know someone is still working on wearables, even if I'm not. This is an attempt to make user interfaces and key-entry out of eyescans.

Keep plugging, hardware and softwear people, I need a wearable so bad.

The Tao of Mac - Python/Grimoire

The Python Grimoire is a fantastic use of wiki for a very very useful resource.

I unfortunately don't use python often enough anymore to be entirely up on the best ways to do every little thing(not that I ever really did), and the Python Grimoire hits a very sweet spot for the occasional experimenter like me, who needs something that works for a problem small enough to do trivially myself, but large enough to need a proper solution.

"The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can be in principle be so precisely described that a machine can be made to simulate it. An attempt will be made to find how to make machines use language, form abstractions and concepts, solve kinds of problems now reserved for humans, and improve themselves." [McCarthy et al. 1955]

It's interesting, to investigate the very roots of a field so varied and strange as AI. At the beginning, when people were just beginning to investigate the possibility of machine intelligence, there seems to have been a much wider and more interesting engagement with the concepts 'in theory'. But their ideas of implementations were like cartoons of even what Weak AI projects do these days. It's a terrible tendency, to generalize and simplify what you don't understand into what you think you do. It leaves you with giant floating concepts which don't actually attach to anything in a functional way.