Ai Dreams Forum

Artificial Intelligence => General AI Discussion => Topic started by: ivan.moony on July 15, 2014, 04:38:22 pm

Title: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: ivan.moony on July 15, 2014, 04:38:22 pm
Here is a site I've stumbled upon some years ago. I just remembered how it impressed me when I've discovered it: http://attempto.ifi.uzh.ch/site/tools/ (http://attempto.ifi.uzh.ch/site/tools/) (click on link: Ace reasoner RACE)

It seems it features deduction and some question answering mechanism. I wonder how would it look like if it was filled with encyclopedic data (like http://babelnet.org/ (http://babelnet.org/) or http://dbpedia.org/About (http://dbpedia.org/About))
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Don Patrick on July 17, 2014, 05:53:25 pm
I can't seem to get the reasoner to work with other axioms, but it is clearly doing a basic inference on the given sample. Expert systems might interest you, they perform inferences on a knowledge database, though I can't say I've seen expert systems around that you can ask questions in normal language.
Interesting knowledge databases. I might use them some day.

If you like this sort of AI, I'm curious about your impression of this one. (http://artistdetective.com/arckon.htm)
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: ivan.moony on July 17, 2014, 06:38:30 pm
I like Arckon very much. I'm impressed by its learning capability. Very good conversationall skills, I can see. Very impressive project, thank you for the link.

However, I would like to see Arckon discovering artificial food, medicaments, new quantum physics formulas and things like that. Also, detailed explanations of how to do things in the world would be nice and I think that it would not be much work from where is Arckon now. It would be just theorem proving on top of translating natural language to logic. I guess we will see it in some future versions. I'm looking forward to it.

Here is another link, not that much impressive like Arckon, but it is an attempt too. Not quite AI, more like a parser + calculator: https://www.cs.utexas.edu/users/novak/cgi/physdemoc.cgi (https://www.cs.utexas.edu/users/novak/cgi/physdemoc.cgi)
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: ivan.moony on July 17, 2014, 06:49:05 pm
Don Patric, please excuse me, I'm just curious, is Arckon a project of yours?
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Art on July 18, 2014, 12:48:27 am
Don / Patrick,

It was nice of you to share a glimpse into some of the workings of your creation. Thank you!

Quite the speed reader he is! ;)

You've obviously spent a great deal of time on this project thus far. What other plans do you have
with regard to additional abilities, refinements, perhaps releasing it, allowing for Internet searches?, etc.

Please enlighten us if you'd be so kind.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Don Patrick on July 18, 2014, 05:06:32 pm
Arckon is indeed my project, hence I am always interested to learn of similar programs that combine deep language processing and reasoning.
I like the calculator, it's simple but thorough :), instead of using grammar, it seems to substitute keywords for mathemetical formula and constants before calculating. Good idea.

Some main features I've scheduled to add are knowledge structure for time, knowledge structure for space, and cause-and-result reasoning, together which will unlock classic problem solving abilities. Getting that up and running will surely take me the better part of next year. I've never had plans to release Arckon to the public, terrorists and mischief makers alike, but I have been contemplating various business models that would allow paying visitors online access to a copy of Arckon, in the event that I run out of funds.
It strikes me that I may have made his output too natural to still be recognisable as the result of a reasoning process (I delibarately omitted some "because"s). It makes sense that people are more easily impressed with superhuman abilities (reading speed) than with a half-adequate human ability (inferences).
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: ivan.moony on July 18, 2014, 05:46:17 pm
Yeah, I have thought of terrorists too in the case of public releasing. But I concluded that police would get a copy too, making a peace balance. But I hate both sides. Using a force from super-intelligent machine scares s**t out of me. When the machine starts systematic search for a problem solution, maybe nothing could stop it from execution. And if reaching the solution takes harming living being, God help us. That's why I think that an intelligent machine should do only things on which all beings agree.

But of course, public wants a gun in its hand for whatever reason they find. And I know they have to be wrong, whatever reason they have. S**t.

However, I'm driven by some crazy thought from which if the world is thriving in its richness there would be no motive for a crime. I'm still hoping for money-less world where everything is free and available. I guess no one wants to make a crime without reason and I would just love to take that reason away from everyone. I hope I'm not just dreaming and that there is a peace of kindness in all living beings, including those terrorists which had a ruff childhood guided by wrong idols.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Don Patrick on July 18, 2014, 07:18:55 pm
Quote from: ivan.moony on July 18, 2014, 05:46:17 pm
When the machine starts systematic search for a problem solution, maybe nothing could stop it from execution.
A kick to the hard disk usually does the trick. Fear is the greatest enemy of all, it is better to replace it with knowledge and care.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Art on July 19, 2014, 01:01:49 am
Do keep in mind that intelligent computer programming isn't restricted to peace loving individuals enjoy harmony in a perfect world. Yeah..."those other guys" know how to program the bad things too! Unfortunately!
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Don Patrick on July 19, 2014, 08:30:11 am
With that regard, I think we should expect trouble in the form of stronger computer viruses before we should expect it from the delicate likes of Asimo. I haven't heard of malicious superviruses for a good while. What I worry about is an arms race for AI, in which safety is disregarded under the pressure, and fear of the other side would cause exactly such an arms race.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: ivan.moony on July 19, 2014, 04:07:30 pm
I believe that autonomous thinking machine should have a suicide button implanted. When it concludes it is dangerous for livings, it should die at once, leaving after mortal note what it died for.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Don Patrick on July 19, 2014, 08:57:52 pm
My program has 4 hardcoded self-destruct sequences built in. Will that do? :)
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Art on July 21, 2014, 02:52:16 am
Only until it decides to modify or rewrite it's code!  ;D
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: ivan.moony on July 21, 2014, 06:55:12 pm
It should make a suicide before changing code.

Anyway, if Arckon is just an answering machine, I think we are safe. The problem would be when the thing imagines some state of the Universe, then systematically tries to reach it, either by using its limbs, or by using words through limbering humans (saying "right" word in the "right" time could cause a lot of mess, I think).
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Don Patrick on July 21, 2014, 07:10:58 pm
Supposing that I would be so careless as to give it programming abilities and leave it unattended, it would likely not be particularly good at programming something as intricate as AI, and rewrite its code into a malfunctioning mess, essentially still self-destructing :)

Yes, Arckon can take no actions other than tell you the answer to your questions, or later, advise solutions to your problems verbally. I may be a mad scientist, but I'm not crazy.
Don't worry, when robots take over the world, it will all be according to my plan, not by accident ;)

The selfdestruct sequences I implemented are actually to prevent it from falling into the wrong hands or being tampered with, a far more likely scenario. Some of the triggers are also encrypted in unconventional ways.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Art on July 22, 2014, 12:42:32 am
<...careless as to give it programming abilities...>

I'm sure you've heard of the scenario of a monkey left alone in a room with a typewriter would eventually, given enough time, compose a novel.
(or something to that effect).

Let's hope that's not the case in your bot. ;)

Actually, it's the way larger mainframe computer and supercomputers that we should be concerned with once they have the ability to self correct or perhaps even construct better, faster code for itself autonomously. Time to move....

Many years a friend and I wrote a program that made use of a simple timer (number of runs) as a safety in case anyone stole our idea. It actually worked too well as we lost track of how many test runs we did and on that last run, it wiped itself clean and destroyed our timer subroutine as well.
Talk about starting from scratch with great amounts of frustration. Live and learn. That was way back in the early computing days shortly after the IBM "clones" appeared on the market or you assembled your own custom box, which we did.
So long ago and we've come so far.

I'm so old that when I was little, there was no color TV. In fact even the rainbows were in Black and White! :2funny:
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Carl2 on July 22, 2014, 02:36:47 pm
  Interesting discussion, I've always liked the idea of a self programing bot,  I believe I first saw that at OTC's site but never got a hold of the programing.  I've learned if you watch black and white TV you will dream in black and white.
Carl2
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Don Patrick on July 22, 2014, 08:30:15 pm
I'm all about control, which is why I have issues with unsupervised AI projects. I compare it to kicking a 3-year old out on the street and expecting it to become a perfectly well adjusted citizen on its own. Shudder the thought of my AI writing novels  ;)
The first scenario I read of this sort of thing involved lego bricks. Small lego robots programmed to build other lego robots who would build other lego robots who would build other lego robots etc etc. - and then the world.

I hadn't thought of limiting the number of runs, that's a good idea :). Did you store the number in a file, or...?

I can't say I know any AI project that is self-reprogramming and not a wanton mess because of it. Though I thought I read something about Google having evolved to a point where its programmers no longer knew how it worked. But I think that's more of a neural net sort of thing, that only improves towards the training data that it is fed.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Art on July 22, 2014, 09:56:02 pm
@ Don,

No, at the time I stored the number within the file / program that was being run, as this was more of a test bed at the time.
I could have placed the number or better encrypted it using an ASCII representation, one line at a time until the program stepped
through all the lines and "added" the ASCII codes together, then a simple conversion and a quick check to see if TIMER = 5 or whatever
then Delete...or something equally as devious.

We once found a "back door" into a (rather primitive by today's standards) BBS online game. Each player used a unique name or Handle. We employed a CHR$(X) + CHR$(Y) +.... trick that added different letters together as the program jumped through various subroutines eventually arriving at our "secret" name which only we knew. The program wrote user's names to a file so that it could recognize new players by their respective names.
All we then had to do was to "log in" under that Secret name and within minutes of playing we were awarded a Flagship for every drone ship we destroyed. Flagships = large points to buy large things. So yeah, it was kind of a trick on our friend who was running the BBS at the time and eventually we told him of our prank. I don't think he would have found the "additions" intermixed within the regular code. It was an Extended Basic language and this was back in the early 1980's so pace yourselves. The logic can still be applied today.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: ivan.moony on July 22, 2014, 11:31:15 pm
Quote from: Don Patrick on July 22, 2014, 08:30:15 pm
I'm all about control, which is why I have issues with unsupervised AI projects.

That should depend on specific algorithm. I'm sure I could work out something you would agree on, if that day comes. Here it is about being in harmony with all living beings, with humans, over animals and bugs to very plants and other uni or multicellular beings. I think unsupervised learning should be allowed if it respects all beings, don't you too?
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Art on July 23, 2014, 10:09:08 am
It is always good and great and wonderful...until someone hacks it. ???
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: Don Patrick on July 23, 2014, 12:45:51 pm
Art: Very interesting. I never learned how to embed variables or code within a program by a program, though I can see how it would be easier in a program that is basically a text file and doesn't require compiling like c++.

Ivan: Most of the time the well-being of beings are in conflict: Humans kill animals for food, so even grocery shopping makes the AI an accomplice to murder. Even if the AI is not hazardous, I just have issues with the lack of wisdom in unsupervised methods. For instance, Google ran an image recognition neural net experiment where they just let the AI loose without goal, and the result was extremely random. Another example: IBM once allowed Watson to learn from the internet, and as a result it kept swearing when it answered. They had to install a specifically designed swear filter.
But if you figure out an ensured way to encode respect for all living beings, I'm interested.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: ivan.moony on July 23, 2014, 01:07:08 pm
The idea is that AI would do only things on which no one objects. If we ask it to kill an animal for food, it should say "Sorry, animal objects, but I can try to find a way to fix artificial food". If we ask it to pass a toy to a kid, it should say "OK, no one objects, I'll do it".

But issues start with kids and drunk people who don't know what is good for them. Their thoughts are contradictory and I still don't know how to resolve this. Usually people know somehow what is better for kids, more than kids themselves. And somehow, I don't like the idea of AI telling others what to think. But maybe something safe would pop up.

There will always be someone who will say: "AI is unsafe, it should make  a suicide". But what would AI do then? Maybe to prove that it is safe, continuing its existance in the world. Maybe proofs are the answer of safety.
Title: Re: ACE Tool - the closest thing to artificial intelligence I've ever seen
Post by: ivan.moony on July 23, 2014, 03:27:11 pm
The following would be a starting point for unsupervised learning. It is an IBNF definition of grammar which, when parsed with all ambiguities, returns an abstract syntax tree (forest more precisely) with all possible interpretations of passed stream of characters.

Code
knowledge {
    abstract word (
        letter {@AnyCharacter \ @WhiteSpace},
        next {@word | @Null}
    ) |
    
    sequence (
        element {
            @word |
            @sequence
        },
        next {(@WhiteSpace, @sequence) | @Null}
    )
}

After parsing, learner code should somehow analyze the forest and induce basic grammar rules. It should find data or elements of sets which are being periodically repeated in each sentence, sequence of words, or in each word (we helped the learner by hardcoding whitespace between words and words could be split in two or more pieces to find morphemes - not shown here). The more data there is to analyze, the more accurate rules would be found by induction.