Artificial Intelligence - Egor - Of Child and Verbs

This shows the basic foundations for Egor, an artificial intelligence brain I am developing. Some more basics - of child relationships and simple verb.

Category:

Science & Technology

Tags:

• semantic
• ai
• parsing
• linguistic
• artificial
• intelligence
• sentence
• grammar
• language
• machine
• learn
• computer
• speech
• english

[waldo2020]

very cool! any code we can look at?

[macdweeb]

No source code or demos in the pipeline for the near future, although I will probably do more videos .. however I am tentatively putting many of the ideas together for an ebook.

This should hopefully be a far easier way of getting across some of the methods (and problems) I have so far come across, and hopefully inspire some more people to explore this area.

[akithered]

Can this system stand the fuzzy logic?

[macdweeb]

The brain knowledge is built with probabilities, so it should support fuzzy logic, although the interpretation side just works in a binary fashion for easier development at this stage. Perhaps more relevant is .. it doesn't just store knowledge, but also the source of this knowledge.

[macdweeb]

So if e.g. john tells egor the cat is black, but jane tells egor that the cat is white, the discrepancy is stored and can be used in giving an answer. And at a later date if egor finds that e.g. john has been feeding dubious information, he can use this in order to come to the best conclusion.

[macdweeb]

It just works with text for now .. you could however use it with any of the many existing speech to text systems pretty easily, there are some pretty good existing solutions for speech to text.

Visual information on the other hand is a whole other kettle of fish, and a huge field in itself which quite a few people are working on. As far as I'm aware though there is no established 'solution' comparable to say, human vision, although there may be some specialized solutions for e.g. driving.

[Esperantanaso]

I love how it responds "OOOKKKEEEeeeeeeyyy."

[mmmAH]

Hmm.. is it only a text processor, or can it actually compute other inputs (ex: video, voice) if given the opportunity?