Brains working towards machine consciousness

  You have IE, bad luck
1 Diagram {x=300}  
2 Content {x=200}  
3 Style {x=400 y=70}  
4 textarea {x=120 y=130}  
5 html {x=270 y=130}  
6 graphic {x=340 y=130}  
7 pic {x=460 y=130}  
8 nodes {y=220 x=100 ln=#FF0000}  
9 links {y=220 x=190}  
10 /body {y=290 x=300}  
11 /html  
- 1.6 has 2.1   
- 1.5 hasA 3.0  
- 2.6 definedIn 4.2  
- 2.4 defaultIn 5.0  {ln=#30E532}
- 3.6 produces 6.2  
- 3.4 canSaveAs 7.0  
- 8.1 definedIn 4.6  
- 9.0 definedIn 4.4  


 

Later, I will ajaxify this, link it up to a server which handles accounts, and thus have a collaborative tool. OK, I know people like drag and dropping things, but this works, and, in theory, is more inclusive.

OK, so this is very silly - I just threw together a little test program (to generate data which might fool a LSA based tool I am planning) which produces a stream of text based on the probabilities of n-tuples drawn from a corpus of multi-authored documents (in this case, various academic papers I have on my hard drive).  This example output is on the basis of 3-tuples (that is, given a 2-tuple, it works on the probabilities of which word follows those two words in the source documents), drawn from a random sample of 40 documents from a corpus of 8,900 (some of which are in German, but