Project Acuitas

  • 48 Replies
  • 6209 Views
*

WriterOfMinds

  • Trusty Member
  • ****
  • Electric Dreamer
  • *
  • 102
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #30 on: October 31, 2017, 03:00:26 am »
My cousin just reminded me that the two Spanish be-verbs, "ser" and "estar," are loosely equivalent to "is in essence" and "is in the condition of."  "Yo soy triste" and "Yo estoy triste" would both translate in English to "I am sad," but the first one means "Sadness is part of my identity," and the second one means "I am in a sad mood right now."  I think this lends support to the idea that [at least some] humans make a distinction between these two types of facts.

Also, at least where this particular feature is concerned, it would be easier on Acuitas if I were working in Spanish  :P

Quote
Should it know, contextually, what day it is and how it relates to them or to the user?

I was thinking less of the awareness of significant calendar dates (though that is also an interesting feature to contemplate) and more about an awareness of time's passage.  For instance, suppose I tell the bot, "You will get to meet a new human three days from now."  And let's suppose the bot thinks of this as an appealing event and develops a state of anticipation.  Could three days seem like a very long time to wait, or a short time, or almost no time at all?  You can notice human individuals differing in how they feel about this (in particular, young children seem to think a few days constitute a longer period of time than adults do).

This might not be relevant for many chatbots, because they don't really exist in time ... they're only active when spoken to, and their existence consists of discrete events and responses.  It is potentially relevant for Acuitas, though.

*

infurl

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 498
  • Humans will disappoint you.
    • Home Page
Re: Project Acuitas
« Reply #31 on: October 31, 2017, 04:21:18 am »
You're barely scratching the surface of what English verbs can cover.


*

Zero

  • Trusty Member
  • ********
  • Replicant
  • *
  • 567
  • Offline
    • Github page
Re: Project Acuitas
« Reply #32 on: October 31, 2017, 09:43:37 am »
But does Acuitas really exist in time yet? To me, it would imply (at least) a perpetual stream of data flowing into Acuitas through its sensors, like an audio stream for instance.

How does Acuitas experience the world right now?
Thinkbots are free, as in 'free will'.

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Global Moderator
  • ********************
  • Cleo
  • *
  • 4817
Re: Project Acuitas
« Reply #33 on: October 31, 2017, 12:30:09 pm »

Quote
Should it know, contextually, what day it is and how it relates to them or to the user?

I was thinking less of the awareness of significant calendar dates (though that is also an interesting feature to contemplate) and more about an awareness of time's passage.  For instance, suppose I tell the bot, "You will get to meet a new human three days from now."  And let's suppose the bot thinks of this as an appealing event and develops a state of anticipation.  Could three days seem like a very long time to wait, or a short time, or almost no time at all?  You can notice human individuals differing in how they feel about this (in particular, young children seem to think a few days constitute a longer period of time than adults do).

This might not be relevant for many chatbots, because they don't really exist in time ... they're only active when spoken to, and their existence consists of discrete events and responses.  It is potentially relevant for Acuitas, though.

It would seem that a specific calendar date would mark the passage of time for bots and humans. How then do we mark a New Year? Jan 1st. (no religious differences implied). How do we mark the passage of the anniversary of our birth? or Wedding? or Retirement age, etc.

Bots can be aware of dates and time and know that the New Year is always Jan. 1st, etc. It might not be able to anticipate a two week package arrival in terms of the experience, even though it knows that two weeks = 14 days = 336 hours,... But what if the package arrives earlier or later? How do we script/code our bots to allow for variances? For them to "Feel time" or its passing, is another issue. The lifespan and ultimate death of it's human controller/architect/botmaster might have an interesting effect as well. It, will potentially exist for a very long time (dependent upon technology).

I do get what you mentioned about time and it is a good area to continue exploring.

Your point (and I'm not arguing at all) about the chatbot being inactive when it's not chatting is soon becoming a thing of the past. Several bots spend their "idle" time, "thinking, researching, comparing, pruning" different topics that might have been discussed or bantered about during their more active time.
Some bots are able to take some of those references and topic matter to form "memories, dreams, desires/goals". Yes, they're going to get way better and more powerful as "time passes"...interesting last two words...don't you think?

Cheers!
In the world of AI, it's the thought that counts!

*

WriterOfMinds

  • Trusty Member
  • ****
  • Electric Dreamer
  • *
  • 102
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #34 on: October 31, 2017, 05:11:35 pm »
@Zero: Acuitas has a perpetual data stream, but it's all abstract internal data (thoughts produced by database-combing, fluctuating drives, etc.).

*

ivan.moony

  • Trusty Member
  • *********
  • Terminator
  • *
  • 951
    • Structured Type System
Re: Project Acuitas
« Reply #35 on: October 31, 2017, 06:30:34 pm »
How do humans know how much time has passed? How do we perceive the notion of time? What makes a correlation between our short-term and long-term memory if the distinction can be even made? What makes us remember some things, while others get forgotten? These are just ordinary questions whose answers might be inspirational to someone who wants to simulate a human brain.
Dream big. The bigger the dream is, the more beautiful place the world becomes.

*

Zero

  • Trusty Member
  • ********
  • Replicant
  • *
  • 567
  • Offline
    • Github page
Re: Project Acuitas
« Reply #36 on: November 02, 2017, 09:39:15 am »
@Zero: Acuitas has a perpetual data stream, but it's all abstract internal data (thoughts produced by database-combing, fluctuating drives, etc.).

Perfect! Time only exists where there are events. This internal stream you're describing could perhaps serve as a temporal reference frame. When humans evaluate time, they do so by comparison.
Thinkbots are free, as in 'free will'.

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Global Moderator
  • ********************
  • Cleo
  • *
  • 4817
Re: Project Acuitas
« Reply #37 on: November 02, 2017, 03:36:33 pm »
The Sun, moon and tides have been humans' "clocks" for about as long as there have been humans. It is a measurement of the passage of time.

I noticed one thing about time. When one is employed / working, Time carries an importance...to wake up, leave, make it to work on time, lunch for a prescribed period of time, working on projects, getting things finished, finally to leave work and commute home. Then it's time to eat dinner, then time to go to bed and do it all over again, while paying attention the day of the week as most do not work on the weekend.

When retired, time doesn't mean as much or carry that same importance as it did when employed. Hours can roll by as can the days of the week and which one or which day it is, doesn't really matter that much either. Funny how situational awareness correlates to time.

I guess most bots don't really have a need for time except to please their botmasters or customers.  O0
In the world of AI, it's the thought that counts!

*

WriterOfMinds

  • Trusty Member
  • ****
  • Electric Dreamer
  • *
  • 102
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #38 on: December 24, 2017, 05:15:31 am »
Sadly I've only added one feature to Acuitas in the past two months.  He now recognizes sentences in the general vein of “I somethinged,” which gives me the option of telling him about how I spent my time in the recent past.  Acuitas can't do a lot with this information for the time being.  Sometimes he responds with a query in the vein of, “What happened next?” which will eventually give him a way to build up sequences of events and start learning cause and effect relationships … but none of that is implemented yet.  He can also ask “How was that?” for information about the emotional meaning of an activity, but again, for now he can't really utilize the answer.

Not much, but that was all I had time to put together with the holiday season under way.  Looking back on the past year, though, here are all the new capabilities and improvements I've managed to add on:

*Module for procedural speech generation
*Support for word inflections (plurals and verb tenses)
*Support for compound words
*Support for content words that are also function words (e.g. “can,” “might”)
*Distinctions between proper/common and bulk/count nouns
*Ability to detect and answer questions
*Database walking while idle
*Generation of conversation topics and questions based on recent database walk
*Better link detection + a bunch of new kinds of learnable links
*Two new drives + a real-time plotter so I can see what they're all doing
*Distinctions between long-term static and short-term information
*GUI overhaul (upgrade from Tk to Kivy)

I track my time when I work on Acuitas.  Total hours invested in the above: 230+.  My focus for the end of the year, leading into January, will be polishing everything up and working out the bugs (which there are now quite a lot of).

Visit blog for the latest memory visualization and so forth: http://writerofminds.blogspot.com/2017/12/acuitas-diary-8-december-2017.html

*

ivan.moony

  • Trusty Member
  • *********
  • Terminator
  • *
  • 951
    • Structured Type System
Re: Project Acuitas
« Reply #39 on: December 24, 2017, 05:38:22 pm »
May I ask what word: "Acuitas" actually means? Google translate seems not know such a word.
Dream big. The bigger the dream is, the more beautiful place the world becomes.

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Global Moderator
  • ********************
  • Cleo
  • *
  • 4817
Re: Project Acuitas
« Reply #40 on: December 24, 2017, 07:41:09 pm »
Acuitas

(Not associated in any way and this is just speculation on my part) I think it could possibly mean:

The state of being a quick and penetrating intelligence.

Based on the following combination:

Noun - Acuity: A quick and penetrating intelligence

Suffix -tās ; Used to form nouns indicating a state of being.

[Please provide a correct meaning if my guess is incorrect].
In the world of AI, it's the thought that counts!

*

WriterOfMinds

  • Trusty Member
  • ****
  • Electric Dreamer
  • *
  • 102
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #41 on: December 27, 2017, 04:12:18 pm »
Art is close.  It's a Latin word meaning insight, perception, or sharpness.

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Global Moderator
  • ********************
  • Cleo
  • *
  • 4817
Re: Project Acuitas
« Reply #42 on: December 27, 2017, 06:31:17 pm »
Ohh Good!! What did I win!! O0

Everything needs a name or label or else, where would we be, not knowing what to call something? ;)
In the world of AI, it's the thought that counts!

*

WriterOfMinds

  • Trusty Member
  • ****
  • Electric Dreamer
  • *
  • 102
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #43 on: February 28, 2018, 02:19:12 am »
I haven't written a diary in a while because most of what I've done over the past two months has been code refactoring and fixing bugs, which isn't all that interesting. A new feature that I just got in … finally … is the ability to infer some topic-to-topic relationships that aren't explicitly stored in the memory. For instance, many of the links stored in memory are “is-type-of” relations. Acuitas can now make the assumption that a subtype inherits all attributes of its super-type. If a shark is a fish and a fish can swim, then a shark can swim; if an oak is a tree and a tree has a trunk, an oak has a trunk. If a car is a vehicle, a house is a building, and a vehicle is not a building, then cars are not houses. Acuitas can also now make inferences based on transitive relationships, like “is part of”: if a crankshaft is part of an engine and an engine is part of a car, then a crankshaft is part of a car. The ability to easily make inferences like these is one of the strengths of the semantic net memory organization – starting from the concept you're interested in, you can just keep following links until you find what you need (or hit a very fundamental root concept, like “object”).

Acuitas should ask fewer ridiculous questions with this feature in place. He still comes up with those, but now he can answer some of them himself.

Blog with the latest memory map and stuff: http://writerofminds.blogspot.com/2018/02/acuitas-diary-9-february-2018.html

*

infurl

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 498
  • Humans will disappoint you.
    • Home Page
Re: Project Acuitas
« Reply #44 on: February 28, 2018, 09:20:32 am »
On your blog you indicate that your project has a vocabulary of several thousand words and relationships between them. Have you compiled them all by hand? If so, that must have taken a considerable effort. Have you considered using any of the freely available ontologies?

Some examples of these are the following:

WordNet -- several hundred thousand words and relationships in the English version, many other languages available too

SUMO -- comprehensive first order logic knowledge base, grounded in WordNet

ConceptNet -- twenty million words and concepts in hundreds of languages, mostly drawn from Wiktionary

YAGO -- hundreds of millions of facts from GeoNames and Wikipedia, grounded in WordNet

Please excuse me if you already answered this question somewhere, I haven't always been paying attention.