QuickAI / News: Recent posts

We go into Alpha with all source required for operation checked in! You need Java and LispWorks (personal version should work). The Java portion requires download of JWNL (Java WordNet Library) 1.3.3 and download and modification of the Stanford Parser 1.6.1. Momentous.

Depending on how you're counting. But am very hopeful that I have much of the theory pinned down, to the place where a general implementation naturally follows. The hard part is doing something simple, to avoid the syndrome of "didn't have time to write you a shorter letter". Still at the day job, so progress is still relatively slow. And there we are.

Working on a general solution, and have been for several months now. However, I've had to take a day job again to pay the bills, so progress is somewhat slower going. Still at it though.

Check out the output in the text file in the doc directory in CVS if you want to see it in current operation. At this point, if you have LispWorks, and you want to see if you can get it working for you, send an email to me and I'll walk you through it (it's complicated).

Have something of a plan to implement a more general logical/inferential capacity. Wish me luck; hopefully it will go smoothly.

Yes, got it to function for basic syllogistic logic. Checked in. A mess right now, but it works!

For anyone who looks here (and has found this in spite of the name change) after much wrestling with theory, we just might be onto something, now. Code is reorganized, and has the beginnings of a new implementation.

I think I've had it right every time that I thought I had it all figured out. Just that the details were not filled in. And here in March of 2009, I finally think that I've begun this project. So the timeline:

I began in march of 2009
I ended in august of 2007
2008 I spent working out detail

Indeed, 2008 I have the one main memory of working on this project, and had very little to show for it, but it was necessary. And now, time flows forward from here. Wish me luck.

http://johnhdoe.com/me/b2.php?m=200901#3201

As we speak, it does some rudimentary "fathoming", mainly connecting relations. The output is still basically unintelligible (to anyone but me), but it does seem to function at a very basic level.

The initialization of the input phrase into machine readable form seems to be working, next to test out the main logic, which has been improved since last I wrote. Hopefully soon some kind of prototype can be assembled, to be tested on OS X and Windows using LispWorks.

Did my first real CVS commit, with a version that compiles but is completely untested (for the most part), and some stubbed. This should delimit the core of Mind 2.0. Perhaps the next commit will add comments, first, while it is fresh on the mind.

Went through major theoretical issues, which caused mountains of headaches and scribbling. I think I finally have a concept of an end-to-end reasoning paradigm, and I am into implementation (and re-implementation, unfortunately). But really, I can't complain.

I have it designed now down to its core functionality, so I have actually begun writing code for this thing. This after 6 years of research. Whew.

Cheers.

I think I have fleshed out the design of the first part, Theoretica.info. I am currently working on a simple implementation of Radiplex Networks by way of creating a spell-checker. All good on all fronts so far.

Deep into design issues at the current time, do have a little code written. Going at it full steam.

Lots of things have been happening to me of late. Let us suffice to say that I've hit the 2 week countdown at my day job. Woohoo! And the rubber is definitely hitting the road. I finally have the concepts I developed in abstraction nailed down into concrete forms that I can actually work with! Life is very, very good.

Have been taking it a little easier as of late, but the project has never been completely out of the foreground. Getting better at understanding how Lisp works (like macros and packages), and am writing up a design going from the general architecture to the detailed design of the thing. Still very optimistic about the whole deal. Perhaps at some point I'll do another revision of the tech spec, but that's not high priority at this point, as it is enough to start from.... read more

Updated the tech spec and spread it out to a few more people. I think I have it at least in my head how this should work, and I have started laying out the design of it, some baby steps. Also been doing some actual coding in Lisp, and I am definitely feeling more comfortable in that. From humble beginnings, like they say...

Finished a VERY rough draft of the tech spec, though I am still figuring some very basic stuff out. My partner has decided not to be a part of this, so it looks like it's just going to be me. I don't suppose it will be that different, as he was not going to learn Lisp anytime soon. Myself, I am getting familiarized to thinking in that language, and on it goes. Email me if you want to see a copy of the specs.

So, what the project will basically entail contains 4 parts. Their names are as follows: Theoretica.info (Theory of Information), Radiplex Networks, G. P. (General Purpose) Reasoner, and ExpertSystem.sh (Expert System Shell). They will define, respectively, a data structure which proposes a new way that information in terms of meaning is conceived, a new paradigm for how information is structured and accessed, what may be called the “intelligence” of this AI, and a practical application for all this smartness in the form of an expert system shell (i.e. such as the venerable program CLIPS).... read more

We've had a meeting of the minds, here at Eonwing, and I've begun the task of working out a technical specification. There's a long road ahead, but really, there has already been a long road here. Wish us luck!