Texai

Only a few usability issues are remaining to be fixed before the online release of the Texai bootstrap English dialog system. Most of them are illustrated by the very first workflow item:

hello stephenreed
* Plural word form for any of these nouns ‘or’?
or (ProperNoun)
OR (Noun)
(y) The plural is ‘ors’.
(n) There is no plural.
(s) Skip this item.
(d) What are the definitions of this word?
Or provide the plural in quotes.
n

You are probably confused by this. My wife and sponsor, BethLynn Maxwell certainly was. Although I plan a help page to explain the dialog format to novice volunteer mentors, it needs to be clear and self-explanatory.

Here is what the workflow item is attempting to communicate.

Continue Reading »

On the AGI-list, this question was asked:

Of the present day used systems, which products do you consider good business cases of applied AI? I’m looking for concrete product examples (not just wide technologies), in either consumer, or business everyday usage, with high profitability, and/or impact factor.

Having read the proceedings and indeed attended several of the annual conferences of the Association for the Advancement of Artificial Intelligence, I easily formed a relevant list of topics to search. Here are the pertinent results from those searches.

• voice recognition products from Dragon
• autonomous industrial robots
• machine vision
• automated scheduling
• constraint satisfaction from Microsoft
• route planning
• symbolic computation
• machine translation
• automatic landing systems
• recommendation systems
• logistics optimization
• stock trading
• pattern recognition
• spam filters

So where is AI currently used in a business context? Obvious answer: many places.

On the AGI-list, Richard Loosemore pointed to the New Scientist article Disorderly genius: How chaos drives the brain. Richard uses one point of the article to advance his theme:

My arguments have always been top-down (complexity has to be there
because the overall features of the system don’t seem to be
implementable without it), so the way I interpret these low-level
results is that the brain has evolved to have all that self organized
criticality down at the bottom BECAUSE that is the easiest way to sort
through all of the different designs, to find ones that give optimum
high level behavior.

But to me, its content confirms an approach that I’ve taken with Texai to tackle AGI complexity.

Continue Reading »

On the AGI-list, Eric Burton said:

What has OpenCyc accomplished that qualifies as step to AGI? Nothing.
The consensus on this list is that OpenCyc should be thrown out: it is
a distraction

During my employment at Cycorp, John DeOliveira and I lobbied successfully for an open edition of the Cyc knowledge base. John and I are still listed as OpenCyc administrators at SourceForge and John not only authored the OpenCyc web site, but subsequently founded the independent Cyc Foundation after leaving Cycorp. Because I was an OpenCyc developer, because I am no longer working for Cycorp, and because I am taking a somewhat different approach with Texai, I believe that I’m well qualified to describe OpenCyc, and argue that its accomplishments are steps toward AGI.

Continue Reading »

Texai needs to keep separate track of what it’s volunteer mentors have taught it, because those facts may be in conflict. I use context to maintain each person’s belief state. There are default contexts that represent the beliefs of Texai. For example, when Texai is taught the plural word form of some English noun, the corresponding new morphological rule object is persisted into a unique context that represents that mentor’s beliefs. Subsequently during parsing an example utterance, morphological rules are selected from the mentor’s belief context first. Only if no applicable rules are found, is the default context searched for applicable rules.  When a sufficient number of mentors agree on the correctness of a particular morphological rule, then that rule is promoted from the mentor’s belief contexts to Texai’s default context

Continue Reading »

Alan Turing was born on June 23, 1912. In 1950 he published his paper Computing Machinery and intelligence in which he proposed his famous test to answer the question “Can machines think?”. As a plausible mechanism for developing a computer capable of thinking he recommended:

Instead of trying to produce a programme to simulate the adult mind, why not rather try to produce one which simulates the child’s? If this were then subjected to an appropriate course of education one would obtain the adult brain.

Turing then elaborated:

We may hope that machines will eventually compete with men in all purely intellectual fields. But which are the best ones to start with? Even this is a difficult decision. Many people think that a very abstract activity, like the playing of chess, would be best. It can also be maintained that it is best to provide the machine with the best sense organs that money can buy, and then teach it to understand and speak English.

The latter is the approach of the Texai project. Lexical knowledge and skills will be acquired so that the system can bootstrap itself into ever increasing capabilities, as taught to it by volunteer mentors. Turing’s requirement for sophisticated sense organs is reduced by (1) a focus on English text rather than English speech, and by (2) the hypothesis that the development of sub-symbolic processing can be postponed and that sufficiently intelligent behavior can be achieved almost entirely by symbolic processing plus spreading activation, which is much simpler and needs dramatically less computing power.

Celebrating this notable date, I expect to release before midnight a online, rudimentary version of the Texai bootstrap dialog, that will enable volunteers to register and to teach Texai plural noun forms.

[UPDATE] I’m postponing the release for a few more days.  Today’s testing revealed that storing each volunteer mentor’s contributions in a separate belief context required small changes throughout the existing code.

On the AGI-list, Lukasz Stafiniak, suggested:

I have another idea you might like less. What about giving the users
the possibility to modify the questions asked, early on? Even not
understood, someone could then try to reparse the questions-answers
later; but to have some mechanisms exposed in a “modify me” source of
the question-originating script.

Actually I like the idea of a volunteer mentor being able to modify the questions asked in a knowledge acquisition script.

Continue Reading »

On the AGI-list, Jim Bromer said:

This [spreading activation] does not provide reasons for any of the decisions made in any way that could be useful to the program. While it might seem reasonable to get the program to learn rudiments of language as a basis from which it would be able to comprehend (some) higher level reasons for decisions, the feasibility of this approach has never been demonstrated in any way.

A design guideline that I’ve followed for Texai is to have cognitively plausible processes when possible. Spreading activation is a well known cognitively plausible process that is described by John Anderson in his book “Atomic Components of Thought” and also at the ACT-R web site. I agree with a weakened version of your statement, namely that spreading activation does not provide deductive reasons for any of the rational decisions made by a Texai agent. Spreading activation is an efficient way to encode contextualized experience in a semantic network, and reinforcement learning can easily be applied to the the links.

Continue Reading »

There is a new download of WordNet 3.0 available which includes a semantic tagging of the glosses.

http://wordnet.princeton.edu/glosstag

With a bit of work adapting this version to the WordNet 2.1 version that is the basis of the Texai lexicon, I should be able to add this tagging information to the Texai lexicon glosses, i.e. to the definitions of each word sense.  The tagging information will allow the Texai parser to unambiguously parse its own word sense definitions, when attempting to understand them.  Given this ability, it should then be possible to automatically gather potentially disambiguating information for each word sense from each discourse context in which it appears.  These automatically gathered contextualized relevance links would supplement, but not replace the need to gather additional ones from volunteer mentors.

In a message, Jason Maskell asked:

if a volunteer gets the noun “spam” – what is he going to be doing? You said you’ve incorporated wordnet, but I’m not sure what you’re doing on your side.

Texai’s current lexical entries for the word ’spam’, derived from WordNet, are the following:

[noun.1 spam]
unwanted e-mail (usually of a commercial nature sent out in bulk)

[noun.2 spam]
a canned meat made largely from pork

[verb.1 spam]
send unwanted or junk e-mail

Continue Reading »