Anyone wants to learn logic?

  • 124 Replies
  • 12264 Views
*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Anyone wants to learn logic?
« Reply #15 on: December 09, 2018, 01:21:56 pm »
Disagree. Watch: If I say "If pigs are blue and metal then I will sleep tonight." yes yes consequence can be true....let's look at the nodes in this sentence:

make up we didn't say pigs...if we do then blue/metal/blue&metal are false (0). Let's say blue/metal/blue & metal are true since fine in the real world as facts.
blue & metal => sleep
1 & 1 => 1
1 => 1
0

Think of these 3 constants as nodes in a hierarchy:
  O
      O
O O  O

In my example, yes the lower 3 are true, but the bigger parse/part is false. Not to say I will sleep is false, but the answer saying if x and v then s is wrong, I don't need it to be so to sleep. Ok it's true but if it means I will suddenly sleep then no false node. It's about validation, not if all are true in last layer. So no, if cause is false then effect can be true - yes, but not this node/sentence sometimes. And also no, if cause is true but effect is false, the top 3rd node again can't be true. We /know/ the lower 3 small parts of speech are true or not, ya, how/why? Cus of validation/proving, but top node 3 must be validated too, and yes it's parts can help out, but it doesn't mean for sure it is true. Again watch - "If pigs sleep then I shower.". [[][]] - [][] are true but [    ] is false.

"If I am blue is false and I am a pig is false then I will sleep tonight."....here we have 2 false leaf nodes and a true 3rd one....the 2 false become (in the processing, not the nodes no they don't change) true by negation, however the final node as whole is false since I don't require them to be false ahem true. Ok as said above same thing, nevermind. It can be true in some cases ex. if I am or not am a blue pig. But not if mean I will suddenly sleep. Time for another: "If I am blue is false and I am a pig is false then I will sleep tonight."......ok this one: "If fire is hot and cows can go to sleep then I will began eating." Here, we need a cause, a true cause, but the right cause (If I am served food I want and am hungry).

"If cows fly is false then I do have more room to fly in the sky."....here the false stays false in memory but is made true for here for now.
« Last Edit: December 09, 2018, 02:05:48 pm by LOCKSUIT »
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Anyone wants to learn logic?
« Reply #16 on: December 09, 2018, 01:24:21 pm »
edited
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Anyone wants to learn logic?
« Reply #17 on: December 09, 2018, 04:58:56 pm »
P.S. I got it now, put it together, now I can start Lesson 3.0....unfortunately there is some changes I made, of course their knowledge is a starting point, I have modified my theory now, integrated it (some I mean, their theory isn't perfect, unless further content changes my mind).
Emergent          https://openai.com/blog/

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Trusty Member
  • **********************
  • Colossus
  • *
  • 5865
Re: Anyone wants to learn logic?
« Reply #18 on: December 10, 2018, 04:25:42 am »
If cows fly then I think I'll just stay indoors! ;)

Seriously Lock, you seem a bit rattled as if trying to find something or some examples that don't fit the proof model or perhaps some other logical area.

Do give things a thorough go around before posting a disagreement and maybe things will become clearer to you.

And is inclusive, Not is exclusive, Or is either, Nor is neither. Sometimes we humans can overthink things.
In the world of AI, it's the thought that counts!

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Anyone wants to learn logic?
« Reply #19 on: December 10, 2018, 05:54:23 am »
"If my computer speaker turned to lava then I would freak out." --- True
s => f
1 => 1
1
The first small fact is false but because of the 'if' it ahem um isn't. Same for saying NOT (negation) ex. "If pigs don't fly then I have more room in the sky to fly my jet."

My theory has it that we only speak true things, at least be believe are true ex. "Earth is flat.". False are not saved, hence no 'negation' is possible. Positive and negative nodes are saved and true. Negation is just nodes saved that are true like "I am not lava." ('not' can be placed almost anywhere). Neutral = false. True nodes have ranks. If you lie to someone and say something false, then you either use or generate a saved false that is sitting alone unrelated in brain, and you only say it when 'good' (reward) is around if you can prove reason to why to say it.
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Anyone wants to learn logic?
« Reply #20 on: December 10, 2018, 06:20:39 am »
see above post too


Indeed, mind boggling!! My todolist aka my artificial Working Memory enhancement was shrinking, now it's double in size cus this course lol! But I'm happy. Now I gotta clean up my todolist after the storm that I solved! My main LTM storage (my notes lol) is in a super huge notepad file. Yesterday wasn't good, played with a GAN for 5 hours, then 8 in my chair on Logic lol.
« Last Edit: December 11, 2018, 05:25:03 am by LOCKSUIT »
Emergent          https://openai.com/blog/

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Anyone wants to learn logic?
« Reply #21 on: December 10, 2018, 12:00:42 pm »
Got any more Stanford courses like this public one?
Emergent          https://openai.com/blog/

*

Art

  • At the end of the game, the King and Pawn go into the same box.
  • Trusty Member
  • **********************
  • Colossus
  • *
  • 5865
Re: Anyone wants to learn logic?
« Reply #22 on: December 10, 2018, 12:38:27 pm »
Not Stanford Lock but MIT courses that you can audit for Free online!
The listing is pretty extensive from cognitive science, pattern recognition, Brain Structure, and its Origins, and so much more!

https://ocw.mit.edu/courses/find-by-topic/#cat=science&subcat=cognitivescience
In the world of AI, it's the thought that counts!

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Anyone wants to learn logic?
« Reply #23 on: December 10, 2018, 01:32:08 pm »
So I take it the courses I checked out only have that small amount of text as the lecture? There is nooo material, despite it says there is full absolutely there....

I go to a page, and have the side menu on left hand side, yes, but each of those pages Online have only some writing.......ain't no book here...........as for book, apparently there is a book they use and I don't have it! The list of its parts shows a amazon link, oh come on! Do I really have to find the books? Now I see pdf links, lots, but they want cash for knowledge!

What gives lol?

And what gives here, no, come on, is this single one course suggesting you read say 9 of these books or more? I take like 2 or more weeks on a single book...or longer sometimes
file:///C:/Users/A/Desktop/9-916-fall-2001/contents/readings/index.htm
« Last Edit: December 11, 2018, 05:22:45 am by LOCKSUIT »
Emergent          https://openai.com/blog/

*

8pla.net

  • Trusty Member
  • ***********
  • Eve
  • *
  • 1299
  • TV News. Pub. UAL (PhD). Robitron Mod. LPC Judge.
    • 8pla.net
Re: Anyone wants to learn logic?
« Reply #24 on: December 10, 2018, 07:12:01 pm »
Ivan,

Would you mind previewing
my new prototype, Emo, here?



      Hello Sir,
 ()_/ Mr. Moony
<\/   I'm Emo.
 []
 ][


Emo is a codename for a simulation
of a physical state for abstract ideas.
My Very Enormous Monster Just Stopped Using Nine

*

ivan.moony

  • Trusty Member
  • ************
  • Bishop
  • *
  • 1722
    • mind-child
Re: Anyone wants to learn logic?
« Reply #25 on: December 10, 2018, 07:24:58 pm »
simulation of a physical state for abstract ideas? What do you have in mind?

*

8pla.net

  • Trusty Member
  • ***********
  • Eve
  • *
  • 1299
  • TV News. Pub. UAL (PhD). Robitron Mod. LPC Judge.
    • 8pla.net
Re: Anyone wants to learn logic?
« Reply #26 on: December 10, 2018, 07:27:32 pm »
Thanks Ivan,

Logic is useful for a smart interface,
before connecting an A.I. engine. Here
is an exclusive preview of my prototype,
code name Emo: 

http://chatbot.ml/emo/

Emo starts learning, using for example:

LEARN the sun IS hot

LEARN an apple IS a fruit

LEARN water IS wet

LEARN planets ARE round

Note: Not case sensitive, "learn A is B"
works, as does "learn Cs are Ds".

However, Emo has no artificial intelligence
engine to apply learning to responses.
So if you ask Emo, "WHAT IS water?" at this
point Emo will not respond, "wet" yet. Emo
stops after learning tuples and placing
them in a table.  That's where you
come in, Ivan, if you wish.

Ivan, please advise, about relational logic,
in terms of using tuples in a smart interface,
before giving the control to an A.I. engine.

Emo has basic emotions. Visual
reactions to fear, comfort, embarrassment,
being focused, bored, happy etc.
Emo has predicate logic depending on
the stimulus, such as to censor curse words.

Emo may show emotions, using for example:

I hurt you

You are boring

You are liked

Focus on it.

In conclusion, Emo is not an advanced system.
Emo is a research system designed to simulate
a physical existence for abstract ideas.

If Emo is down, that is just me programming.
Just wait a minute or two, and Emo will wake up
again for you.

The goal is not to become an A.I. but to prepare
the stimulus for an A.I. to get a head start in the
interface.



My Very Enormous Monster Just Stopped Using Nine

*

8pla.net

  • Trusty Member
  • ***********
  • Eve
  • *
  • 1299
  • TV News. Pub. UAL (PhD). Robitron Mod. LPC Judge.
    • 8pla.net
Re: Anyone wants to learn logic?
« Reply #27 on: December 10, 2018, 09:59:50 pm »
Training blocked output temporarily

However, "fish = animal" was learned.

Notice how it prepends the tuple

instead of appending it.

Thanks!
My Very Enormous Monster Just Stopped Using Nine

*

ivan.moony

  • Trusty Member
  • ************
  • Bishop
  • *
  • 1722
    • mind-child
Re: Anyone wants to learn logic?
« Reply #28 on: December 10, 2018, 10:50:10 pm »
8pla.net, i'm afraid you caught upon a huge task. Natural language processing and logic isn't developed that much to be capable to implement the complete version just like that. Even if you learn the proposed Stanford material, there is still a big chance you'll respond to it like "This stuff is unfinished!", or "If it takes this much to learn this fragment, how much does it take to learn, or invent the whole system?" You know that story - the more we know, the more we realize we don't know. But I think the the story extends to - the more we know, the less we are certain it is correct. Right now, the science has a fragment in a form of: propositional calculus, predicate logic and higher order logic. But it wasn't enough (at least it was that much abstract), so mathematicians recently started to invent topological space theory, category theory, and probably some more ambitious areas I'm not yet aware of. Scientists are even writing a collaborative open source book of knowledge theory that anyone can contribute to. You should see all that mess lost in a vast of complicated scientific papers, it almost seems like it never ends.

The point is: no one currently has a capability to answer how to connect a natural language (I propose Treebanks) piece to reasoning logic (whatever form it takes, logic, lambda calculus, type theory, ??? ). If you are inventive enough, you could be the one to connect all the pieces, but the question is how much time you are willing to spend. A few weeks? A month? A year? Two? Ten? The whole life? It is hard to compose something meaningful in a short time, but it is not impossible if you are lucky, at least luckier than I was. Some perfect scenario would be: (1) learn a bit - program it and show it; (2) learn more - program more and show it; (3) invent atop of all that - program it and show it; (4) ...; (X) conclude AGI - program it and enjoy. That would be the perfect scenario, but I wasn't much lucky to extrapolate the checkpoints to do something meaningful. Some curse was all the way above me, blasting and smudging the checkpoints to unrecognisability, forcing me to reach the final AGI, without letting me to have a satisfaction to implement the fragments on the way. I simply didn't have anything that worked as a fragment, but was only a piece of a broken whole that can't be programmed because it is contradictory. Finally, I gave up, pinning myself to the ground. However, I still hope - not to reach AGI - but to walk out of all this mess with something that could be used in the real world. I'm hoping for a scientific language based on logic. I'm trying to get out of all I learned by now, and it still isn't as trivial as I could expect after all these years of reading scientific papers and having more or less delusional thoughts about improvement I should be able to accomplish on the way.

I'm  not complaining (I'm glad I tried to do something in my life), I'm just exposing the status of my knowledge - a lots of fragments, I can do this and that (mostly manipulate formatted scientific knowledge), but I have no idea how to round more or less pieces into a meaningful whole other than the language I'm developing.

That being said, I'm probably the last person that should direct the development of Emo. But that doesn't mean I couldn't help with answers. But to do than, I need more specific questions. Try to gather your thoughts, place imperatives, ask yourself what exactly do you want Emo to do, how much time you want to spend, and then pose your questions. I'll be around with my theoretical logic knowledge, and I'll be happy to answer anything that's in my humble mind.

Specifics, please. Otherwise I'm as lost as an atom in the Universe.

For a start, could you please specify what do you want Emo to be able to do, if I didn't scare you off with the dragon named Logic?

[Edit]
I just thought of something I posted here some time ago. Maybe you'll be interested in element-subset-set reasoning fragment. I think it takes the time to read the Stanford material from this thread (you have to know the resolution method), plus some few weeks to implement the fragment in Javascript, or something. Sorry, I've got nothing simpler for now.
« Last Edit: December 10, 2018, 11:21:30 pm by ivan.moony »

*

8pla.net

  • Trusty Member
  • ***********
  • Eve
  • *
  • 1299
  • TV News. Pub. UAL (PhD). Robitron Mod. LPC Judge.
    • 8pla.net
Re: Anyone wants to learn logic?
« Reply #29 on: December 11, 2018, 12:50:09 am »
Ivan

I converted your markup language to source code to test the results:


STEP 1. Markup

<x> is a <X>.
<x> is a <A> or <x> is a <B>.
<x> is a <A> and <x> is not a <B>.
If <x> is a <A> and <x> is a <B> then <x> is a <C>.

STEP 2. Code

$x is a $X.
$x is a $A or $x is a $B.
$x is a $A and $x is not a $B.
If $x is a $A and $x is a $B then $x is a $C.

STEP 3. Data

$x="Ivan";
$X="genius";
$A="teacher";
$B="student";
$C="brainiac";

RESULTS:

Ivan is a genius.
Ivan is a teacher or Ivan is a student.
Ivan is a teacher and Ivan is not a student.
If Ivan is a teacher and Ivan is a student then Ivan is a brainiac.1

____________________________________________________________________________
1 An exceptionally intelligent person.
My Very Enormous Monster Just Stopped Using Nine

 


Say good-bye to GPUs...
by MikeB (AI News )
March 23, 2024, 09:23:52 am
OpenAI Speech-to-Speech Reasoning Demo
by MikeB (AI News )
March 15, 2024, 08:14:02 am
Google Bard report
by ivan.moony (AI News )
February 14, 2024, 04:42:23 pm
Elon Musk's xAI Grok Chatbot
by MikeB (AI News )
December 11, 2023, 06:26:33 am
Nvidia Hype
by 8pla.net (AI News )
December 06, 2023, 10:04:52 pm
How will the OpenAI CEO being Fired affect ChatGPT?
by 8pla.net (AI News )
December 06, 2023, 09:54:25 pm
Independent AI sovereignties
by WriterOfMinds (AI News )
November 08, 2023, 04:51:21 am
LLaMA2 Meta's chatbot released
by 8pla.net (AI News )
October 18, 2023, 11:41:21 pm

Users Online

287 Guests, 1 User
Users active in past 15 minutes:
squarebear
[Trusty Member]

Most Online Today: 312. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles