|
| View previous topic :: View next topic |
| Author |
|
|
Mind.html
|
Joined: 05 Jan 2007
Posts: 6
Location: Cyberspace
|
|
Posted: Feb 08, 2007 7:13am Post subject: Forth for Modular AI |
|
|
As we each contribute to the Modular AI Project
by coding artificial intelligence in our chosen
programming language to create artificial minds,
Mind.Forth serves as an example of a mature but
primitive AI consisting of well-defined modules.
http://mind.sourceforge.net/aisteps.html is a
list of suggested mind-modules and the sequence
in which they need to be created, so that the
AI program constantly maintains functionality
as each new mind-module is coded and debugged.
Mod-AI participants for any given programming
language should each keep an entire copy of
the Mod-AI codebase in the target language,
so that any module being coded is seen to
work properly with the rest of the AI code.
AI-evolution-wise, it is actually a good thing
if the separate copies of the original AI code
evolve and mutate into diverging streams of
genetic differentiation and divergence, as long
as the human coders share their most advanced
work based on their own specialised expertise.
At each stage of coding the modular AI, even
from the very beginning, an AI User Manual
should be maintained alongside each separate
Web-hosting of the developing AI code base,
similar to the Mind.Forth AI User Manual.
AI coders may eventually notice that the
academic AI establishment begins to publish
details of the emerging modular AI program.
The Association for Computing Machinery (ACM),
for instance, has published Mind.Forth papers:
ACM Sigplan Notices 33(12):25-31 in 1998
Mind.Forth: Thoughts on AI and Forth
and ACM Sigplan Notices 39(12):11-16 in 2004
Forth and AI Revisited: BRAIN.FORTH.
Early participants in our Modular AI Project
may become hot candidates for consulting roles
in the emerging AI industry or for outright
hiring as proven AI programmers, based on
the functioning AI minds pioneered here. |
|
|
|
|
Martin
|
Joined: 05 Jan 2007
Posts: 84
Location: England, worcester
|
|
Posted: Feb 08, 2007 1:02pm Post subject: |
|
|
|
Hmm, well I could program some of those modules - but they'd all be in visual basic |
|
|
|
|
MessageForums.net
|
Sponsored Links
Posts: 177
Location: Earth
|
|
Posted: Today 9:21pm Post subject: |
|
|
|
|
|
|
|
|
FJRAIMind
|
Joined: 19 Feb 2007
Posts: 2
|
|
Posted: Feb 19, 2007 4:41am Post subject: Forth for Modular AI |
|
|
Greeting all.
An online associate pointed me to this site and I thought I needed to be here.
I have been fascinated with Forth for a very long time. Sometime last year I found Mind.html in Forth posted on the web and began following it. I soon found myself adding to the program but using W32Forth.
Most of the work I have done is in the area of I/O and restructuring the mind core arrays. I have done little to nothing to the basic logic of the core. I am leaving this to the author (ATM). I was interested in visual stability and efficiency of the processes and usage of computer memory. I have added an Internet access interface, where the AI can upload to its own website an updated webpage. Actually it can FTP any web site given the address, user, & pw info. Also is capable of downloading from any website files or web pages.
I order to better share what I am doing I put my work on its on website and I update the link regularly. AIMind-I.com. If the program is running the page is update with in- process info every 5 minutes.
Why FORTH? I see it as possible the AI mind could generate its own set of sub-functions / processes building on the Forth Library, besides FORTH is portable. I do not have to have it installed on any computer. I carry it along with me on a memory stick. This is especially handy when using a computer not your own, like at work or for that matter at the library.
I have noticed the more I run the program the more I go back and add to the basic mind library of words. At the present there are ~ 185 words consisting of nouns, verbs, pronouns, adjectives, number, symbols, conjunctions, adverbs and names. The program is not able at this time to use adjectives, adverbs, conjunctions, numbers or symbols but if I use them as input to the running mind it is able to identify them correctly. Also is able to identify via parsing new adjectives and adverbs provided I follow simple rules;
adjectives ending in FUL, LESS, BLE, IC, ICAL, ING or LY for adverb.
POS: 1=adj 2=adv 3=conj 4=interj 5=noun 6=prep 7=pron 8=verb 9=names
10=affirmation 11=Symbol 12=numbers 15=Exec
Gender: Male=M=77 Female=F=70 Neuter=N=78
Tense Plural=P=80 Singular=S=83
Oh yea my use of a plural noun is also identified and is associated with the singular tense provided I use the singular form first. But the mind does not have the ability just yet to associate plural nouns with appropriate verbs. I have also attached a gender flag to all words, M, F, N, eventually to be able to associate gender with concept words.
Why do this at all? Only one reason – its fun. _________________ Frank J. Russo |
|
|
|
|
TrentIllician
|
Administrator

Joined: 05 Jan 2007
Posts: 44
Location: USA
|
|
Posted: Feb 19, 2007 11:56pm Post subject: |
|
|
Well Frank, welcome to our little group, as of now, as I'm sure you've noticed, we're still having a little difficulty getting 100% organized, but it's coming together. I look forward to our continued working together.
- Trent |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You cannot download files in this forum
|
|