Google Blogoscoped

Forum

Memecodes

Andreas [PersonRank 0]

Wednesday, November 17, 2004
19 years ago

Philipp,

What's the current status of the Memecodes Evolutionary Page Experiment? How many pages have spawned offspring, what are the most popular queries? Did you think about implementing a stats page (or actually did implement one)?

I didn't know about this project until you posted the url in the seraphim proudleduck thread. I think it's a great idea, and I'll publish a short article on the project at www.aiplayground.org in the next days. Do you have any other online experiments I might not know about?

Philipp Lenssen [PersonRank 10]

19 years ago #

Thanks for the idea for a Memecodes statistic. I add one to the front-page now.

Philipp Lenssen [PersonRank 10]

19 years ago #

I don't think I have more experiments of this sort exactly. Though of course there's a lot of fun stuff to be found in the Archive
http://blogoscoped.com/calendar/

Philipp Lenssen [PersonRank 10]

19 years ago #

Memecodes Statistics:
http://blogoscoped.com/archive/2004_11_17_index.html#110072328811290305

Andreas [PersonRank 0]

19 years ago #

It's interesting to see the experiment in work:
http://memecodes.outer-court.com/6574.html
"My name: #6574 – My birthday: 2004-11-17 20:50:22
My parents: site%3Amemecodes.outer-court.com + #648"
That's what happens when you (for some other reason) activate referrers for five minutes.

How did you create the initial pages? Doesn't look like random words from a dictionary, did you use the Google API?

Has the total number of pages decreased or increased? How many pages are still "alive"?

Philipp Lenssen [PersonRank 10]

19 years ago #

I created the initial words by using a small dictionary (27,000 words), randomization, and several methods which I thought would make it look more like natural language.
For example I have a set of punctuation characters (! : ; and so on) as well as a set of common "glue words" (the, of, and, to, a, in, for, ...). I also add words from a set of well-known brands (Apple, Coca-Cola, Google.com, Ikea, Nike, ...) and celebrities (Edward Norton, Edward Platt, Edward Rigby, Edward S Brophy, ...).

As for "dying" pages, my idea was to let pages die naturally because they are pushed to the end of the page, and then out of the page. I display 2150 Memecodes "citizens" at any time on the front-page. Also, the page is around 150 K, and Google only indexes around 100 K, so I suspect it will slowly start to ignore older Memecodes.

Forum home

Advertisement

 
Blog  |  Forum     more >> Archive | Feed | Google's blogs | About
Advertisement

 

This site unofficially covers Google™ and more with some rights reserved. Join our forum!