All this week I’ve talked about AI in theory, but why not chat with an actual AI right now? Cleverbot can hold a conversation pretty well – just type in the box and hit Enter, and it’ll “type” a response. For instance, a second ago we started a conversation (no joke!) just like this:
Me: Do you like my blog?
Cleverbot: No.
Gulp. Well, at least it’s honest. Anyway, Cleverbot isn’t actually intelligent (as you’ll discover if you try to talk about the same thing for more than a few lines) but it’s fun for a while.
Programs like Cleverbot are called chat bots. There’s an annual prize for the best chat bot, called the Loebner Prize. Want to win $5,000? Write your own chat bot and win the contest!
Not exactly artificial intelligence, but artificial life: Conway’s Game of Life is a very cool, very simple game invented back in 1970. You can’t lose or win, but you can make some amazing patterns. If you’ve never tried it before, give it a shot. You’re in for a treat.
I can’t let AI Week pass without linking to Isaac Asimov’s short story, “The Last Question.” Not only the best AI story I know, this is the best short story I’ve ever read, period. It’s a quick read.
Did you know there’s a whole organization dedicated to studying and preparing for the Singularity? Welcome to the world of the Singularity Institute.
Speaking of the Singularity…I’ve posted this link once before, but it’s so perfect I have to put it up again. This comic from SMBC demonstrates the Singularity in four simple, elegant pictures. Don’t know if that’s what he intended or not, but that’s certainly how I interpret it.
Don’t forget, it’s AI week at the Trube blog too! He has his own thoughts on language and intelligence, the Singularity, and the AI from Deus Ex. He also posted a forty-minute story of his own, and today he’s going to talk about the best Star Trek episodes with Data. Check it out!
Got any links to share? Put ’em in the comments! Have a stellar weekend, and see you on Monday.
Brian D. Buckley 
Oooh can’t believe I let the week go by without sending you my pet AI theory…
The oldest attempts to create artificial intelligence are Legal Systems.
Even though the operating system is painfully slow and embarrassingly low-tech, the general idea is to ask a question and get a sensible answer. Perhaps even an answer that a human is not capable of.
The law is even rudimentarily self-aware.
I’ve run this past lawyers and they tend to like it. How does it sit with AI people?
I like it! This actually happens with all sorts of human systems: corporations, research groups, armies, open source development projects. The group does things, and knows things, that are beyond the reach of any single member.
In fact, this is very much how my own AI design operates. Lots of unintelligent “agents” communicate with each other, and form an intelligence through their mutual efforts.