From: Eliezer S. Yudkowsky (firstname.lastname@example.org)
Date: Tue Jun 05 2001 - 00:11:25 MDT
The following message is sent on behalf of Ben Goertzel.
-- -- -- -- --
Eliezer S. Yudkowsky http://intelligence.org/
Research Fellow, Singularity Institute for Artificial Intelligence
> So what about this mysterious software system I envision?
> Basically, I want to encode all knowledge and processes in a
> single system and represent it all graphically (2D / 3D
> whatever). Imagine visually sorting a linked list and storing
> relevant information about memory and speed characteristics
> (O(n) etc.) of the algorithm. Then building in layers of
> abstraction to allow all sorts of metaphors (functional
> isomorphisms) between heterogenous systems (all the while
> optimizing as necessary to beat the costs of abstraction).
> Now wouldn't that be a lot prettier and more fun than nasty old
> Mizar syntax? If Eli would just put out a version of Flare,
> maybe I could start working on that graphical front end :) When
> you want to understand a beautiful algorithm, isn't the best
> way (for humans) is to see / play with a graphical animation?
Let me humbly suggest a place to start here
We created a knowledge representation language at Webmind Inc. called
It's not that different from some other KR languages, but I think it's
natural and usable...
Processes can also be expressed in KNOW, though the initial emphasis in
developing the language was on declarative knowledge. A process is
representable as a particular kind of higher-order relation.
The attached spec describes KNOW. The representation of processes in
of KNOW is currently not described except in proprietary documents, but it
could easily be extracted from them.
A KNOW visualizer would be a nice place to start, in my view...
[The attached KNOW spec overflowed a Majordomo limit, and has been changed
to a URL pointing to Ben's site. The main content was reposted at Ben's
request. -- Eliezer]
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:36 MDT