TribeStorming

Begin at The Beginning

What's the cliche about the journey of a thousand miles? It starts with one small step. Begin at the beginning, as they say in wonderland. Continue until you get to the end. Then stop.

This ain't wonderland. It's the world, and for a whole bunch of us, it's the web. There's no beginning to begin from. Just plug and play.

There is beginners mind. Take a deep breath. Clear the mind. Ahhh.

Start somewhere.

Eliza

I mentioned last issue that I've dug up an old copy of an "artificial intelligence" computer psychologist program called Eliza and ported it to Javascript. The idea of Eliza mystified me when I first read about it as a little boy. A computer program capable of carrying on a conversation! Wow! But now that I've seen it, I know Eliza's mostly smoke and mirrors.

Just for fun, I posted to the Javascript newsgroup and asked for suggestions. You can talk to eliza at http://www.manifestation.com/neurotoys/eliza.htm

... But don't even bother unless you have a version 4 browser. I spent forever wading through Netscape's Javascript reference, and even though I followed its examples, it turns out that there's a little bit of my code doesn't run on old versions of Javascript.

I learned a lot from all this:

Source code documentation really sucks.
Eliza, as I found her, really sucks.
You can meet some cool people on the internet.

George

I didn't get much of a response from the Javascript newsgroup. Two people wrote me back. The first guy was the one who pointed out that it didn't work in version 3 browsers. The second guy rewrote it.

George Dunlop has taught me a thing or two about software lately. About taking an idea and running with it. He took my little port of Eliza and rewrote it completely. You can talk to his version on his website ( http://www.peccavi.com ) and see the difference. He's still tinkering with it, as far as I know.

What's amazing to me is this. I wanted a javascript version of Eliza for my website. There were no javascript versions. I found a stripped down pascal version, and stripped that down a little as I ported it. Now there was a retarded javascript Eliza. I casually mention this on the internet, and somebody swoops down and hands me a good javascript Eliza for my website. Why? Because he's having fun with it!

Think With Your Tribe

I come up with so many ideas that I can't keep up with them. I have notebook after notebook of them. That's why I decided last issue to put them all on the web. Make them searchable. I've been playing with that. (There's more to Manifest Station than meets the eye! A lot of stuff is in development, just behind the scenes...)

Newsgroups fascinate me. A bunch of people with common interests. A tribe. You can post an idea, and other people will think about it for you, at their leisure, when they want to. You only have to say something once, yet so many people give you feedback. What a deal!

What a way to create things! Not just Eliza. Look at GNU. Linux. Perl. Half the internet was built by people just having fun. Heck, even netscape's publishing the source for it's next browser!

Information Explosion

The world's huge. I can't learn everything. I don't want to. But when I want a simple piece of information, by God I want it now.

I know how to program. I don't want to go look through a book. Or a 50 page website with no search features to tell me how to generate a random number in Javascript. Only to find out that the feature doesn't even work on most browsers so you've got to write your own routine.

I love reading. I checked out 10 books on artificial intelligence. I went through most of them in an hour each. Most of them overlapped. Large chunks were irrelevant.

That's why we have hypertext. The point isn't to keep someone on your site as long as possible. That's a short-sighted commercialist viewpoint. The point is to make it easy to learn. Easy to think.

Hats off to the Robot Wisdom Pages. (http://www.mcs.net/~jorn/home.html) They taught me this.

First Hint of Tumboodo

I talked last issue about an idea drive. A search engine for knowledge. Not web pages or unfiltered postings. True, useful knowledge.

I started with code. I like to pick on Netscape's javascript references, but the truth is, I've never seen anyone that had what I wanted. I want to say, "I know how to generate a random number in BASIC. Just tell me how to do it in C." And WHAM. There's a line of code in C. Or Java. Or Perl. Or Lisp. Whatever.

So I started building something to do that. It'll work. And it'll work for any kind of knowledge. More next issue.

Machines to Think With

I'm finding more and more that software is a tool to think with.

I use emacs. It's a text editor. It's the most amazing tool I've ever found for clarifying my thoughts. It lets you write collapsible outlines. It can run lisp programs. It even comes packaged with a version of Eliza! It can chop up a document, and recombine the words:

Think With en playing with that's amazing to Think With
What a way to cript Hats off to true, useful knowledge. A bunchable. tarted with code. Internet, and miles?
Sublishing the sourcenes...
Newsgroups for any kines. It can your site as little retarded javing fun. Why? Becaught me thoughts the enginey of and even up dug Manifest Station.com

If only James Joyce were here to see this!

Thought Bots

I've come to believe more and more that machines should help us think. Not just artificial intelligence in computers, but augmented intelligence in humans.

I picture my website as a tool to help me think. Not point and click, but enter an idea and get presented with similar ideas that I've had before, or that other people have suggested, perhaps months in the past. A free-association engine!

That is, a website could be a conversation partner. Except by talking to it, you'd talk to the collected and compiled wisdom of the people of Earth. Maybe we could find something useful to do with it. :)

Bot Thoughts

I'm entering the Loebner competition next January. The winner is the one who produces the most humanlike aritificially intelligent program.

It's a Turing test. Human judges, human players, computer players. All the players try to convince the judge they're humans. All the convincing's done over computer terminals. If in the course of the conversation, the computer manages to convince the judges its a human, the computer's programmer wins. If no one does it, then the most human computer gets the prize. No one's ever done it. Yet.

To find out how to make a stupid computer act like a human, a lot of people ask how we wound up making a stupid mammal act like a human. It's not inherent, you know. Look at the stories of people raised by wolves. I don't think they'd have much chance at the Turing test. There's more to humanity than the hardware.

Finale

Someone once suggested to me that the better a person's fantasy life, the better his real life. I've bought into that. You know, the better you make the world inside your head, the better the world outside becomes.

I've been remodelling my brain long enough. Now I've found in software a way to share my fantasy world with the rest of civilization. I have a feeling my brain's going to get remodelled all over again.

That's okay. I'm up for it.

HTB Lab Notes

Whew! That was fun.

Hey, I'm employed for real. I design databases and back ends for websites. How fortunate, considering my interests.

Today is Lori's birthday. Happy birthday, sweetheart! I Love You. :)

I don't know when I'll decide to write the next issue. It'll be a surprise. Meanwhile, if you liked this one, pass it on.