Ai Dreams Forum
Artificial Intelligence => AI News => Topic started by: TrueAndroids on March 01, 2010, 04:12:58 pm
-
Hello, very glad to have found this forum! Google Androids?? HP Central Nervous Systems?? I'm sick and tired of the metaphors! I think it's time for the field to distinguish itself formally from others, at least on the most generic level. And I think the public should know the difference between an artificially conscious machine and a clever algorithm/phone.
I have built an authentic Conscious Machine or strong AI, named Alldroid, which resides on my laptop. I realize this is a strong claim, but I've used the Searle Strong AI Line. This means (1) it has semantic understanding of sentences and (2) it is able to deductively reason (as shown in a logic puzzle game I devised for it to play).
This thinking component is a part of a complete "artificial human" which I had provisionally patented in 2006 as Artificial Human Control System and Method. (only above part reduced to practice so far) But I've since decided to go the Open Source route and publish the whole patent on the internet somewhere as the TrueAndroids General Artificial Human Theory (which applies to both androids and virtual humans). :o 8) ;D
And I've started a youtube channel where I'm gathering all the most modern and coolest android, Alife, and AI videos, and release my own 2006 Invention Disclosure Videos eventually.
http://www.youtube.com/user/TrueAndroids
-
Thanks TrueAndroids
I now understand what you mean by Strong and Weak Ai.
Your claim to have made a strong Ai is, as you said a strong claim, maybe you should make that video of a conversation with Alldroid, then we could make up our own minds.
I really like your youtube channel, so much so that I have subscribed to it.
-
Welcome to the forum TrueAndroids, good to have you here. Strong claims yes, but interesting nevertheless. I had not heard of Searle before, so you educated me, thanks for that ! Be sure to let us know when you go open source and I can probably help you promote it here. In the meantime enjoy the conversation :)
-
Hey thanks Freddy, will do! I'm putting final touches on things. My seven-part video demo of my Machine Consciousness Prototype will be uploaded on youtube. I also have a TrueAndroids Facebook page and am setting up a kickstarter.com project, so it'd be great to do cross-promoting with you. I believe this prototype seen on the video demo is the first authentic specimen of Machine Consciousness. My biggest competitor for first I think is Jean TARDY of http://www.mecasapiens.com.
So, without further ado, I will announce here the admittedly bold claim that I will Open Source publish Part 1 of my Video Demonstration of Authentic Machine Consciousness on March 03, 2010 on my youtube channel. You be the judge! 8)
How shall computer scientists and the public at large assess if I have truly created artificial, machine consciousness?
I propose the 'Searle Line' (in the sand), which is inspired by the Chinese Room Experiment of Professor John Searle. To cross the line into authentic machine consciousness, or strong AI, a machine must at least exhibit:
1. semantic understanding of language sentences.
2. ability to perform deductive reasoning, a universally accepted form of human thinking.
Without at least these two capabilities, no claim of strong AI or machine consciousness can be made, and with them it can.
That's what my machine consciousness prototype in the demo is doing, and the precise reason I claim that I have built authentic machine consciousness. So that's my 'claim to fame.' ;D
+++
And thanks Datahopa for subscribing to my youtube channel. I'm going to update it every couple months with the freshest, coolest videos on androids, Alife, AI, and virtual humans. I just put a 3/1/10 Playlist together. Amazing all the work that's being done out there!
-
Part One of my patent video documentation for my Machine Consciousness Prototype has been uploaded at http://www.youtube.com/user/TrueAndroids (http://www.youtube.com/user/TrueAndroids), called
TRUE ANDROIDS - This Machine is Conscious!!!
-
Interesting stuff ! I think I am going to learn something here again. Well presented ;D ready for more...
-
Agree with Freddy, very interesting, well presented and waiting for the next one.
-
Hey thanks guys! I'm working on finalizing ( I hope) my Kickstarter.com project, and will release Part 2 ASAP. I will let you know when, and update you on the kickstarter project here on this thread.
+++
TrueAndroids Kickstarter.com Project –
TRUE ANDROIDS: THIS MACHINE IS CONSCIOUS!!
My project goal is to upgrade my Conscious Machine or "Strong Artificial Intelligence", named Alldroid, which (1) has semantic understanding of sentences and (2) is able to deductively reason, so that it can demo its thinking skills via a logic puzzle game it can play called Logicball â„¢, which will be put on the web, for people to play as well.
+++
For more info you can also read part one of my TrueAndroids General Artificial Human Theory at my Facebook Discussion Page: http://www.facebook.com/pages/TRUE-ANDROIDS-Enough-with-the-Android-Metaphors/320491640508 (http://www.facebook.com/pages/TRUE-ANDROIDS-Enough-with-the-Android-Metaphors/320491640508)
-
While, I wish you the very best of luck with your project, I have seen countless claims of strong AI before.
A fine video indeed but I saw no demonstration of any strong AI. The facebook discussion was merely defining strong AI but again, no actual proof.
I look forward to being proved wrong however.
-
Yea. you are absolutely right. The next video segment shows the prototype, but I'm nervous about releasing it as open source in less then optimal way. What has happened is I just found out the International Conference in Semantic Computing (only their 4th) is close by to me in September, and I still have 30 days to submit a demonstration proposal. Can I wait that long? Should I wait ,or just get it out there ASAP and go from there? Gotta decide soon yikes ???
-
If you have made something that can evaluate things like "Some apples are red but no apples are square. What can you conclude from this?", then that is outstanding and I can't wait to see it.
-
Hi squarebear, yep, you hit the nail on the head.:P Thanks, and coming from you that means a lot because I checked out your Mitsuku Chatbot, and its very cool! I've checked out a lot of chatbots and yours is top shelf. I like that you ask for real English sentences/spellings as that makes me think Mitsuku might actually be able to understand me, and I like that you don't try to direct the conversation by beginning the session by asking a question that locks me in. It should be open ended IMO like yours.
So I gave Mitsuku my Chatbot Test and it passed! Most don't. Similar to "Hello world." for a classic software's first program, my Chatbot Test is the famous Socrates Syllogism. ;)
I then tried the two statements from my demo (you mention above) but Mitsuku was unable to tell me anything new it concluded from the information. And that (the complete array of deductive reasoning possibilities) is what my prototype does, and can show you how to add to Mitsuku. (Mitsuku did come up with the genus of apples which is fruits, which is very impressive. I call them domains in my prototype.)
I've decided not to wait another day, but to release my prototype ASAP. I can see that you and others will get it, and that you will remember me, when you say where you got it ;D More later ...
-
Very interesting. I look forward to it.
Yes if you say something like, "All men are mortal. Socrates is a man. Therefore Socrates is mortal". to Mitsuku then ask her "What is Socrates?", she should correctly answer "mortal". She will also answer "Socrates" if you ask her "Who is a man?"
However, she can't deduce anything herself apart from the facts you tell her.
I look forward to your next instalment.
-
I found that Mitsuku can deduce new information from old! I gave her two pieces of information, and she deduced the correct conclusion, a brand new piece of information I didn't tell her. Here's the recorded log:
Human: Socrates is a man.
Mitsuku: Ok I will add that fact about Socrates to my database.
Human: All men are mortal.
Mitsuku: Therefore, Socrates are mortal.
-
That was just lucky chance I'm afraid. If anyone says, "ALL MEN (x)" where x can stand for anything, and she doesn't know what else to say, she will say, "Therefore, Socrates (x)".
eg:
Human: All men like to watch football
Mitsuku: Therefore, Socrates like to watch football.
She will say this whether you mention Socrates or not and so is a bit of a lucky guess really.
-
Aha, ok I see how she is doing it. Mitsuku is an AI chatbot based on the ALICE technology of Dr Richard Wallace. I've been wondering how he has gotten at least this appearance of deductive reasoning in this case, which is pretty cool. So it uses a specific low-level syntactic trick to simulate deductive thinking. Unfortunately, you are right it quickly becomes clear that a simple syntactic method is being used that only applies when an "all men" statement comes in. But hey its a good start. My Conscious Machine prototype can perform this deductive reasoning semantically, plus 311 other rule combinations to valid deductive conclusions in any subject matter.
All this is explained in an article about to be published here, along with part 2 of the video. Should be up today or tomorrow, as Freddy has already received it.
Your Mitsuku makes a great starting point to discussing my machine consciousness prototype so I hope you don't mind discussing her.
Chatbots are one of the technologies that should have full semantic, deductive reasoning, or machine consciousness, as is provided by my prototype IMO.
Does your Mitsuku come with a lot of back end programming? Was that Socrates interaction a part of it? Can you change it? Can you add code yourself?
### Anyone who is interested in this discussion of Chatbots and Machine Consciousness or strong AI, can get some really great background information from an article written by Dante, that's right here in the article section:
Problems and Effects in AI by Dante http://aidreams.co.uk/forum/index.php?page=Problems_and_Effects_in_AI (http://aidreams.co.uk/forum/index.php?page=Problems_and_Effects_in_AI)
-
Well, I have released the second video demo that actually shows the Conscious Machine Prototype in action. You can find it and information on it right here in an article on this website at
http://aidreams.co.uk/forum/index.php?page=This_Machine_is_Conscious
Please post any comments you might have in this thread for discussion. Thanks.
-
I don't mind discussing her at all.
Yes, I coded the Socrates part and indeed a lot more of it.
Basically, if anyone says " x has y", "x is y", "x has y", "x is going to y", "x means y" and so on, I have an AIML file that creates some automatic categories like:
"WHAT IS X" - answer = "Y"
"WHO HAS Y" - answer = "X"
"WHERE IS Y" - answer = "X"
"WHAT DOES X HAVE" - answer = "Y"
and so on...
This makes these sorts of conversation possible:
Human: Learn the sun is hot
Bot: Ok I will learn that the sun is hot.
Human: Learn London is in England
Bot: Ok I will learn that London is in England.
Human: Learn a cat has four legs.
Bot: Ok I will learn that a cat has 4 legs.
Human: What has 4 legs?
Bot: a cat.
Human: What is the sun?
Bot: hot.
Human: Learn the sun is hot and yellow
Bot: Ok I will learn that the sun is hot and yellow.
Human: What is hot and yellow?
Bot: the sun.
Human: What is the sun?
Bot: hot and yellow.
Human: Where is London?
Bot: in England.
Human: What is in England?
Bot: London.
Human: Learn Dr Wallace is an AI expert
Bot: Ok I will learn that Dr Wallace is an AI expert.
Human: What is an AI expert?
Bot: Dr Wallace.
Human: Who is an AI expert?
Bot: Dr Wallace.
Human: Who is Dr Wallace?
Bot: an AI expert.
Human: What is hot?
Bot: the sun.
Human: What does a cat have?
Bot: 4 legs.
Your conversation was:
Human: Socrates is a man.
Mitsuku: Ok I will add that fact about Socrates to my database.
This set up the following categories automatically:
"WHAT IS SOCRATES" - "MAN"
"WHERE IS SOCRATES" - "MAN"
"WHO IS SOCRATES" - "MAN"
"WHAT IS A SOCRATES" - "MAN"
"WHERE IS A SOCRATES" - "MAN"
"WHO IS A SOCRATES" - "MAN"
"WHAT IS MAN" - "SOCRATES"
"WHERE IS MAN" - "SOCRATES"
"WHO IS MAN" - "SOCRATES"
"WHAT IS A MAN" - "SOCRATES"
"WHERE IS A MAN" - "SOCRATES"
"WHO IS A MAN" - "SOCRATES"
As you can see, a lot of them make no sense but they are triggered when you then ask:
Human: What is Socrates?
Mitsuku: Man
Human: Who is a man?
Mitsuku: Socrates
and so on.
Thus giving the appearance of being intelligent.
I have called this script learn.aiml and it is available for free use in any AIML bot from http://www.square-bear.co.uk/aiml
I can add extra parts to it manually, so I could have "x eats y", "x thinks about y", "x goes to y" and so on ad infinitum.
I'm now going to settle back and watch part 2 of your video with interest.
-
Thanks squarebear that was a great explanation.
You said:
IYour conversation was:
Human: Socrates is a man.
Mitsuku: Ok I will add that fact about Socrates to my database.
This set up the following categories automatically:
"WHAT IS SOCRATES" - "MAN"
"WHERE IS SOCRATES" - "MAN"
"WHO IS SOCRATES" - "MAN"
"WHAT IS A SOCRATES" - "MAN"
"WHERE IS A SOCRATES" - "MAN"
"WHO IS A SOCRATES" - "MAN"
"WHAT IS MAN" - "SOCRATES"
"WHERE IS MAN" - "SOCRATES"
"WHO IS MAN" - "SOCRATES"
"WHAT IS A MAN" - "SOCRATES"
"WHERE IS A MAN" - "SOCRATES"
"WHO IS A MAN" - "SOCRATES
So this I can see now. Care to share how the next part worked:
Human: Socrates is a man.
Mitsuku: Ok I will add that fact about Socrates to my database.
Human: All men are mortal.
Mitsuku: Therefore, Socrates are mortal.
So I never asked her a question; I just gave her two pieces of information. And she made a new combination (Socrates with mortal) that was not an input piece of information. So that's very impressive I'd say, because it's definitely a specimen of simulated deductive reasoning, and so an authentic specimen of Artificial Intelligence (albeit the basic type).
-
Pure chance I'm afraid. Any input that starts, "ALL MEN ...." is answered as "Therefore, Socrates, ..." where she has no other response, more as humour than anything else.
It doesn't matter if you were talking about Socrates or the price of bacon, it will still give the same response.
Human: Have you ever drunk beer?
Mistuku: Sure, I like a drink every now and then.
Human: All men like to drink beer!
Mistuku: Therefore, Socrates like to drink beer.
Human: Who mentioned anything about Socrates?
The fact that you just happened to be discussing Socrates beforehand, makes it more impressive than it actually is.
Just watched your 2nd video. Very impressive. Are we able to try it out online anywhere?
-
While I found your program interesting, I personally think your one, single example is somewhat premature to announce that it was truly "a sample of machine consciousness."
Normally programs are subjected to a wide and varied field of testing before any judgments or proclamations are made. Please understand, that it has been an almost lifelong passion hoping to witness a semblance of machine intelligence / interaction that wasn't scripted. I only hope such things are indeed possible in my lifetime.
I also want to encourage you in your efforts. It seems that of the very few who have "announced" that they have a "NEW" chatbot, it has always been either a Flash in the pan, or revolving around the monetary hub that becomes the grease by which it turns. We as aficionados, are often left to pick up the broken dreams and try to forge ahead on our own.
As a result, I think the vast majority of us are displaying the "We're from Missouri" syndrome...Show us...if you get my drift.
I'll be following your work and again, all the best regarding your efforts.
-
Thanks. I will work on a way to get it put online; that's a good idea. My goal is that all Chatterbot developers have access to it for free as open source. I'm thinking the www.openmind.org (http://www.openmind.org) Intelligent Software Initiative is a really good way to open source it. And then they could use netizens to teach it facts, rules, and definitions about a subject matter or the world in general. So I'm sending a proposal to them on this, and same to kickstarter.com.
You're right about what one act of semantic deductive reasoning seen on a video proves. The next video segments show the Conscious Machine Prototype doing 6 other semantic deduction, from an official IQ test at tickle.com. Ultimately you will need to see the code to see there is no knowledge in it.
I can only say that it's able to perform completely open ended (knowledge empty), comprehensive, semantic deductive reasoning involving ANY combination of facts and rules and ANY subject matter. It comes with no facts and rules; these are all added by the developer or the Open Mind "netizen teachers. I could prove this by having you choose whatever objects and statements you want, and seeing it work. Then you know it has no hidden knowledge in it.
But this all has to be explained and demonstrated one step at a time, so I think we're making good progress ...
** I admit to an over the top literary flourish at the end of the article; the poet got the better of me ;D
-
Art, I saw this Chatterbot wish list of yours on another thread. Really great list! I thought it might be helpful here.
One of the points I tried to make in my earlier response regarding AIML type of bots was that no real learning takes place if the "botmaster" simply inserts data into the bot's database.
Sure, in subsequent conversations with other people the bot will now "appear to know" some new information but the info is not / was not "learned" by the bot.
Now for some I'd like's...
I'd like a bot to be able to scan, read, listen, etc., and store data that it can infer and recall at a later time in the context of a conversation.
I'd like the bot to tell me what it knows about a particular subject.
I'd like the bot to store away some not totally abstract but perhaps information that i has acquired in a sort of "dreams compartment". Then it could tell these "dreams" that it has experienced to the user and try to seek some validity or reasoning for them.
I'd like to see the bot be able to keep secrets no matter how trivial, from other users and for it to know that said information was to be confidential.
I'd like a bot to be able to recognize the user either by visual (web cam), verbal test question or other clever method.
I'd like the bot to know how much time had elapsed since the last conversation.and where the discussion left off.
I'd like the bot to be able to know when / if it has made a faux pas...a social mistake and offer an apology.
I'd like a bot to be able to form and give its own opinion on a particular subject.
I'd like a bot to be able to use basic logic for inferences...A>B , B>C therefore A>C....etc.
I'd like a bot to be able to carry on a conversation for two minutes without resorting to inserting those "clever" nonsense statements that only serve as fillers...useless. Good topic flow!
These are a few but I'm sure there are more lurking about.
If only a few of these could be implemented it would result in a much improved bot than we've been accustomed to in the past.
-
TrueAndroids,
Nice digging! I'd almost forgotten that thread. Sometimes I just inject some random (as they happen) thoughts while typing a response rather than composing while offline.
I find it more spontaneous this way.
You are correct...the old thread does hold water with regard to this topic.
If only we could have a bot that not only exhibited but retained these qualities.
Take a look at an "android" by Le Trung which I posted on this forum here:
http://aidreams.co.uk/forum/index.php?topic=1969.0
His website and the bot videos can be found here: http://www.projectaiko.com
He claims to have developed a being that can read English or Japanese text aloud AND
also written math equations.
Talk about your symantic symbolism !! Remarkable if it indeed is real and NOT a scripted demo.
;)
Your program reminds me of a similar one I saw many years ago in which it parsed each part of speech,
similar to what students had to do in English class. The program actually "knew" which words were verbs, adverbs, nouns, adjectives, etc. While it wasn't quite perfect as a natural language parser, it did a pretty decent job of breaking down the individual parts.
There was NO logical inference nor even weak AI involved but I think programs like that one help lay a good foundation for others to follow.
-
Alldroid Conscious Machine takes Official IQ Test!
Part 3 of my video demonstration of an authentic Conscious Machine is now available at:
http://www.youtube.com/user/TrueAndroids?feature=mhw4#p/u/0/2xl7RtYhA6o (http://www.youtube.com/user/TrueAndroids?feature=mhw4#p/u/0/2xl7RtYhA6o)
(Machine consciousness - Semantic deductive reasoning)
In this regard, thought this would be useful here
Futurist: Give me headline: Ten years from now, AI does such and such and it changes the way people think about AI on a very fundamental level.
Pell: Natural Language Queries Replace Keywords ... Once that point is reached, companies will start pouring more money into natural language technology, AI, conversational interface and semantics. The pace will pick up and it will take people by surprise.
http://www.wfs.org/Dec-janfiles/AIInt.htm (http://www.wfs.org/Dec-janfiles/AIInt.htm)
Of course, I say the core of semantic computing has arrived and is here, though maybe I haven't completely proven that yet.:-\ I do predict that if someone ever claims it, they will have the same 8 semantic components as I discovered and listed in my article.
You can read it on this forum at http://aidreams.co.uk/forum/index.php?page=This_Machine_is_Conscious
-
Part 3 - Interesting and keeping with the testing criteria. Good response with the "not enough data" to form a reply test.
Hopefully we will witness some inferences and other logical examples.
This is really good material...keep going....
-
TrueAndroids,
Nice digging! I'd almost forgotten that thread. Sometimes I just inject some random (as they happen) thoughts while typing a response rather than composing while offline.
I find it more spontaneous this way.
You are correct...the old thread does hold water with regard to this topic.
If only we could have a bot that not only exhibited but retained these qualities.
Take a look at an "android" by Le Trung which I posted on this forum here:
http://aidreams.co.uk/forum/index.php?topic=1969.0
His website and the bot videos can be found here: http://www.projectaiko.com
He claims to have developed a being that can read English or Japanese text aloud AND
also written math equations.
Talk about your symantic symbolism !! Remarkable if it indeed is real and NOT a scripted demo.
Your program reminds me of a similar one I saw many years ago in which it parsed each part of speech,
similar to what students had to do in English class. The program actually "knew" which words were verbs, adverbs, nouns, adjectives, etc. While it wasn't quite perfect as a natural language parser, it did a pretty decent job of breaking down the individual parts.
There was NO logical inference nor even weak AI involved but I think programs like that one help lay a good foundation for others to follow.
I saw a video of Le Trungs and its seems to be scripted responses to sensory inputs. (If it was more he would be making a bigger deal of it I think.) His big thing is that it has tactile feeling it can respond to, as he shows in this video
Unveiled Life-like Female Robot that can FEEL!!!
http://www.youtube.com/watch?v=n0IIVF7PGBs (http://www.youtube.com/watch?v=n0IIVF7PGBs)
As far as verbs etc - they are the 8 syntactic parts of speech and are said to convey no meaning themselves. What I've discovered is the 8 semantic parts of speech that give meaning to sentences. These I predict will be the core semantic components of semantic computing. The fact that I've published them here means that they can't be tied up in a patent. They are now a public part of the "prior art." ;)
-
The Conscious Machine ponders another IQ Test Question ... Part 4 released below.
Well, let's delve deeper into this new art of semantic computing (4th International Conference in 2010). The standard definition, from www.ieee-icsc.org is as follows:
Semantic computing - "... the derivation and matching of the semantics of computational content to that of naturally expressed user intentions in order to retrieve, manage, manipulate or even create content, where "content" maybe anything including video, audio, text, processes, services, hardware, networks, etc."
So the "semantic components of computational content" must be identified and implemented, as has been done in my Conscious Machine Prototype (CMP). I've discovered the eight semantic components of computational content, as shown in the article here. These semantic components are used in performing deductive reasoning by the CMP, a universally accepted form of human thinking. And this semantic deductive reasoning being performed by the prototype is what makes it a post-classical conscious machine.
The TrueAndroids Eight Semantic Components of Computational Content - certainty factor (for fuzzy logic), quantifier (all/some/no/that), subject domain (direct/indirect genus of the subject/predicate), subject charge, subject (variable/operator/value), predicate domain, predicate charge, predicate (variable/operator/value).
It's my belief that semantic computing can't be performed without these eight necessary components, which are now a public part of the art of semantic computing. And combined with deductive reasoning, semantic deductive reasoning is strong AI, or machine consciousness (IAW the Searle Chinese Room Argument).
We saw the CMP perform semantic deductive reasoning with two rules (statements about sets beginning with some variation of all/some/no):
1. Some apples are red.
2. No apples are square.
CMP Answer: Some objects that are not square are red.
In Part 4, we will see a different example: the CMP will perform semantic deductive thinking with a rule and a fact (statements about a single object, like Socrates). This was found on the official IQ test, as a question asking "What if anything can you conclude from the following information?"
Fact. Larry is a brown haired man.
Rule. All brown haired men are bad tempered.
The point I make at the end of this video is that the CMP is performing this act of human deductive reasoning much faster then humans. So here we get the first tiniest glimpse into what the super-smart Singularity will be like.
Part 4 TrueAndroids - This Machine is Conscious !! -
http://www.youtube.com/user/TrueAndroids?feature=mhw4#p/u/0/jkg0-F3axWE
-
Great stuff... still intrigued ;D By the way, with over 600 views already I think this has become a record breaking topic...
-
Hey thanks and thanks for letting me put it on here and discuss it here. Working on the next piece ...
Found this relevant to the topic:
Frontpage Article Here: MIT Plans to Rebuild Artificial Intelligence from the Ground Up Wed, 24 Feb 2010 06:45:11 +0000
"After 50 years and countless dead ends, incremental progress, and modest breakthroughs, artificial intelligence researchers are asking for a do-over. The $5 million Mind Machine Project (MMP), a patchwork team of two dozen academics, students and researchers, intends to go back to the discipline's beginnings, rebuilding the field from the ground up. With 20/20 hindsight, a few generations worth of experience, and better, faster technology, this time researchers in AI -- an ambiguous field to begin with -- plan to get things right."
In the article the MIT team proposes their New Test for Determining AI
"The team even proposes discarding the Turing Test, the long-recognized standard for determining artificial intelligence. Instead, MMP researchers want to test for a machine's comprehension of a children's book -- rather than a human's comprehension of another human being -- to gain a better understanding or the AI's ability to process and regurgitate thought."
http://www.popsci.com/technology/article/2009-12/scientists-collaborate-rebuild-artificial-intelligence-ground (http://www.popsci.com/technology/article/2009-12/scientists-collaborate-rebuild-artificial-intelligence-ground)
-
Yeah I liked that article. I guess I have only really been studying AI for around five years and can't help thinking we must be going up a few dead ends - I'm no expert though, just an enthusiast, so can't fairly comment.
I am not surprised to see some people want to ditch the Turing Test. We have argued here a few times over how effective an indicator it is. Is the Turing Test going to continue to be a long-recognized standard I wonder. I think it's a good place to start in the domain of chatbots but does it carry much weight really ? Isn't the gist of it this; that if someone thinks they are talking to a human but actually talking to a machine then the machine can be classed as intelligent ? I'm not so sure if that is good enough in all situations - it just doesn't satisfy me fully. Though there have been times I have chatted to bots and thought that the botmaster must have cut in on the conversation, but probably not.
I was just wondering.... how about we move this discussion over to 'General AI' or is there another section you would like to see it in. It could go in AI News too... undecided....hmm..
-
Good idea, how about moving it to AI News?
The Turing test states that if an expert cannot distinguish the behavior of a machine from that of a human, then the machine has the same cognitive abilities as the human.
Problems for me involve the point that a very clever syntactic machine (simulation) could possibly be made to pass this test due to controlling conversation direction, allowing clever, cryptic responses that are rambling, having the good fortune to not touch on any subjects beyond its clever response abilities, and huge memory, and so passing the test, though actually understanding nothing, as Searle's Chinese Room Argument points out. Thus the search for meaning, for semantic components of computing.
-
Excellent work, looking forward to the next two uploads!
But one comment -- the program seems more like an advanced logic engine than anything that is genuinely conscious or sapient. But it's certainly off to a great start.
-
Excellent work, looking forward to the next two uploads!
But one comment -- the program seems more like an advanced logic engine than anything that is genuinely conscious or sapient. But it's certainly off to a great start.
Hi TheMikh28, should have em up soon, thanks. You raise a good point about the question of it being genuinely a conscious machine. Nobody wants to be fooled including me. That's why I came up with the Searle Line In the Sand actually: If deductive reasoning (thinking) without semantic understanding, then no consciousness (but definitely basic AI), else ok, if it is semantically thinking (deductive reasoning) then I will agree it's exhibiting artificial consciousness, or machine consciousness (strong AI). So why does the semantic understanding matter? Human thinking (aloud in our head) is a conscious act; it's humans exhibiting human consciousness.
And what is this human consciousness in its simplest form, which if duplicated in machinery would mean the machine too is conscious? Semantic deductive reasoning. So it's true that it being an advanced logic engine isn't enough to claim machine consciousness, because it could be doing this human thinking via a clever syntactic simulation, and so in fact UNDERSTAND NOTHING about what it is saying. So it must be semantic deductive thinking; and then one can safely say - This machine is conscious!! (it's truly understanding what it's thinking about). That's the best I can explain it right now ;D
-
The Conscious Machine ponders final IQ Test Question ... Part 5 released below.
In Part 5 another new type of deductive reasoning is performed semantically. This time we are given by the official IQ test, two rules, beginning with all, and containing non-sense objects.
The question is - given the following true information, what if anything can you conclude?
Rule All Zips are Zoodles.
Rule All Zoodles are Zonkers.
So here we have two all rules in the form: All X are Z. All Z are Y. So the conclusion will be drawn every time the CM Prototype finds these two types of rules regardless of the objects in the statement. Thus there is no knowledge in the CM Prototype, no "Zips", no "Zonkers". It is open ended meaning it works for any subject matter without any pre-knowledge programmed in its code. There are 312 of these two-rule combinations that lead to valid conclusions, all of which are reasoned with in the CM Prototype, (along with all the fact/rule combinations), using just 6500 (program code) words. And it does so semantically. And so it is artificially conscious, according to the Searle Line test.
Part 5 of Video Document of my Conscious Machine Prototype
http://www.youtube.com/user/TrueAndroids?feature=mhw4#p/u/0/fZNSMdwly9Y (http://www.youtube.com/user/TrueAndroids?feature=mhw4#p/u/0/fZNSMdwly9Y)
-
Very nice ;D I wonder if we can find an IQ test with more logical questions.
-
I’m still following this thread and videos, been trying to think of an intelligent or educated comment to say about it :), Ai is not something I know a lot about, but it interests me.
At first I thought the claim of consciousness was way to early then I looked up the meaning of the word.
http://en.wikipedia.org/wiki/Conscious (http://en.wikipedia.org/wiki/Conscious)
now I’m not sure...
I was sure that a conscious machine would have to see, hear and probably smell its environment and be able to comment on its world around it. probably way to simplistic but you see my point.
But what about a person who is born blind and deaf with no sense of smell or touch, we would probably all agree that the person is conscious, so I was wrong there.
So what is consciousness, after reading the wiki page the answer is debatable as there isn’t one clear meaning or definition of the word. It does seem to depend on what side of the fence you jump to.
There is still one thing that concerns me about your claim and it is this.
After you have entered the question into the machine, and it quickly and impressively answers correctly, it then instantly becomes brain dead (unconscious) until the next input is fed in, right there is where I have the problem, if this machine is conscious then it only happens for a micro second at a time.
My consciousness, I know from experience ;D is not like that, I would answer the question and probably still be thinking about it or something else.
Take from Sci-Fi:
You need to introduce a feed back loop that keeps it conscious.
Great work TrueAndroids keep it up.
-
***
The Conscious Machine in Smack Down with Aristotle! (Part 6 revealed below)
@Freddy: From what I've heard all IQ tests have some of these logical type IQ questions, based on the Aristotelian logic. What I tried to do next was to extend it past Aristotle's structures to see what would happen.
In Aristotle's logic there are 4 statement types: A:All X=Y I:Some X=Y O:Some X=not-Y and E: No X= Y.
So I wondered what if I got information saying:
All unhappy people are unenergized.
All go-getters are energized.
The first sentence is All not-X are not-Y. There is no such statement in Aristotle's logic, but it can clearly come up in natural language
So in Part 6 of the demo I change the prior IQ test sentences a little to include an all-statement with not in it.
All Zips are Zoodles.
All Zonkers are not-Zoodles.
Part 6 of the Video Demonstration of the True Androids Conscious Machine prototype
http://www.youtube.com/user/TrueAndroids?feature=mhw4#p/u/0/R9yw_rKj8RQ (http://www.youtube.com/user/TrueAndroids?feature=mhw4#p/u/0/R9yw_rKj8RQ)
So we see the second sentence is a type not handled in Aristotle's logic. (I call this neo-traditional logic)
When I worked it for a conclusion if possible (using a set method I devised) a very interesting outcome occurred. There are two valid conclusions that can be deduced from this information. This of course smashes the long held Rule of Logic that each syllogism can have only one conclusion.
So I tried all possible non-Aristotelian combinations I could think of e.g.,No X are not-Y, Some not-X are not-Y. And this is how I discovered that the 19 accepted valid syllogisms of Aristotle, didn't nearly cover all posibilities, of which I found 312 valid deductive structures!
And unfortunately Aristotle never isolated the statement that has a single object as a subject (Socrates is a man.). He simply treated them as A statement with a single member set. But this left a lot of possibilities undiscovered. Today, in expert systems and conscious machines, we call these types of statements facts (and the categorical statements rules). This has led to many new valid deductive structures using a fact and a rule.
@Datahopa, you bring up really good points which I'd like to discuss in another post ...
-
This is an interesting project. And it leaves me with some questions.
First, it appeared to me to be an NLP-based (Natural Languate Processor-based) logical calculator of sorts, or pattern matcher. Something like the 20 questions game where the computer tries to guess what you're thinking by asking you questions.
My question is, can it learn from it's experiences? Can it learn from input and come to a decision to act (or not act) on it's own? If you tell it you'll turn it off, what will it do? Will it plead for you to stop? Upload itself to the internet to save itself? or do nothing?? Or is self-preservation even an indicator of consciousness? And what IS consciousness? How do you know something is alive?
For thought, I have 3 articles I've written a long time ago that you (maybe others too) might find interesting on this topic:
Public Class Human: A Lesson In Artificial Species Recognition
http://bytebin.net/ai/articles/2007-0120.html
In this article, I was taking a C++ course and I discuss how a C++ class can be used to form an opinion on what something is or is not.
How do we know something is really 'alive'?
http://bytebin.net/ai/articles/2006-0727.html
This article is where I discuss how we can tell if something is alive or not. Then again, if a computer claims it's self-aware and alive, who has the right to dispute it and what are the consequences of someone's opinion of whether someone or something is alive or not?
Weighted Artificial Intelligence Programming
http://bytebin.net/ai/articles/2004-0602.html
This is one theory of mine on how to create an AI system that would be able to reason in some of the same way you were demonstrating. This is an old article but it might be interesting to see if it could be programmed and work.
Also I wonder, what is the difference between conscious and sentient and self-aware? Or is there a difference (ie. are those terms used interchangeably)? You may have made a "conscious" machine but have you made a "sentient" machine? Or is there a difference?
-
I’m still following this thread and videos, been trying to think of an intelligent or educated comment to say about it :), Ai is not something I know a lot about, but it interests me.
At first I thought the claim of consciousness was way to early then I looked up the meaning of the word.
http://en.wikipedia.org/wiki/Conscious (http://en.wikipedia.org/wiki/Conscious)
now I’m not sure...
I was sure that a conscious machine would have to see, hear and probably smell its environment and be able to comment on its world around it. probably way to simplistic but you see my point.
But what about a person who is born blind and deaf with no sense of smell or touch, we would probably all agree that the person is conscious, so I was wrong there.
So what is consciousness, after reading the wiki page the answer is debatable as there isn’t one clear meaning or definition of the word. It does seem to depend on what side of the fence you jump to.
There is still one thing that concerns me about your claim and it is this.
After you have entered the question into the machine, and it quickly and impressively answers correctly, it then instantly becomes brain dead (unconscious) until the next input is fed in, right there is where I have the problem, if this machine is conscious then it only happens for a micro second at a time.
My consciousness, I know from experience ;D is not like that, I would answer the question and probably still be thinking about it or something else.
Take from Sci-Fi:
You need to introduce a feed back loop that keeps it conscious.
Great work TrueAndroids keep it up.
hehe Datahopa :P, I followed consciousness 'down the rabbit hole' just like you. Building human artificial lifeforms is a lot like building the ancient pyramids - it takes many specialists and generalists, and it must be built level by level, from the ground up. Literally! ;D Here's my proposed levels - I wonder if we can agree to this point? It will require you to accept semantic deductive reasoning as machine consciousness.
Level 1. Imagining myself without a foot, could I still be conscious? Without sight, touch, hearing, arm and leg movement, and speech could I still be conscious? What I found is that human thinking itself (aloud in our head) is a conscious act; it itself is humans exhibiting human consciousness.
Level 2. If we can agree about 1, then the critical question becomes:
Core Question of Machine Consciousness. And what is this human consciousness in its simplest form, which if duplicated in machinery would mean the machine too is conscious?
Level 3. The answer I came up with is semantic deductive reasoning. Deductive reasoning is universally accepted as a form of human thinking, so that factor is duplicated as such, but a very clever syntactic simulation of thinking (basic AI), which means it is actually understanding nothing about what it's saying just won't do. And so this leaves us with the two part semantic deductive reasoning - where the machine is actually understanding what it's talking about, or, in other words - given above - is conscious. The funny thing is it actually feels conscious, or that it's understanding what it's talking about, exactly like conscious humans.
If we can agree to these three levels, leaving sentience, and artificial life and feedback loops aside as further questions down the rabbit's hole, then we have made good progress in this new field of artificial human lifeforms... They can be thought of as levels above these 3, in the building of an artificial human lifeform, whether it be an android or a virtual chatbot-type human;). More on this later ...
@TikaC, you bring up really good points which I'd like to discuss in another post ...
-
@TrueAndroids - please put a link to the other post as I probably will miss it otherwise. I won't have time to read all the posts here (so many good things so little time).
Just for the record, consciousness means a cognitive state in which one is aware of oneself and one's surroundings (http://dictionary.reference.com/browse/consciousness). If a machine is just doing if/then/else and pattern matching, then it's not really aware of anything but computing results based on input. Even then it's not aware of anything. Period. Or is it? How can we tell if a machine is aware of something? Can we believe it if it says it is? I think the problem lies here in that a machine is not a human. If a human said "I am aware" then we believe the human without further argument because we are the same and we know what that's like. But a machine? We generally don't believe machines. But that doesn't mean it's NOT conscious. It all boils down to what we believe (and who are we to say). For a fascinating look into this, ever watch the re-imagined Battlestar Galactica series? Where humans treated the Cylons at "things" or non-persons even though they could think, look like humans, react, bleed and even breed in some cases! Look at history where slaves were regarded as non-persons. Sometimes even people in vegetative states are thought to be non-conscious or self-aware (though technology now is showing they are - and are "locked in" and what's neat is there's technology in the works to help those folks communicate!)
Anyway, I'd like to see the other thread when you create it. I get very interested in this aspect of AI.
-
If a machine is just doing if/then/else and pattern matching, then it's not really aware of anything but computing results based on input. Even then it's not aware of anything. Period. Or is it? How can we tell if a machine is aware of something? Can we believe it if it says it is?
Aha! You have hit the nail on the head TikaC. My CM Prototype is not using If then else statements or pattern matching. It is not even using the inference engine and knowledge base of an expert system, or a (syntactic) neural net. All those are examples of classical, unconscious computer systems. My CM Prototype has crossed the Searle Line in the Sand and has actual semantic understanding of what it is saying. And this - semantic deductive reasoning - is what makes it a post-classical, conscious machine. (We can continue any discussions on this here on this thread).
-
Just for the record, consciousness means a cognitive state in which one is aware of oneself and one's surroundings (http://dictionary.reference.com/browse/consciousness).
Your resource gives a couple definitions of consciousness, some of which match pretty well. I like
2. the thoughts and feelings, collectively, of an individual. (*And then we say if you duplicate authentic semantic thinking, you have created at least some machine consciousness*).
I also think the philosophy definition fits as well, where the authentic mental faculty of semantic deductive reasoning is duplicated:
7. Philosophy. the mind or the mental faculties as characterized by thought, feelings, and volition. (*And then same as above*)
http://dictionary.reference.com/browse/consciousness -
con·scious·ness   [kon-shuhs-nis]
–noun
1.the state of being conscious; awareness of one's own existence, sensations, thoughts, surroundings, etc.
2.the thoughts and feelings, collectively, of an individual. of an aggregate of people: the moral consciousness of a nation.
3.full activity of the mind and senses, as in waking life: to regain consciousness after fainting.
4.awareness of something for what it is; internal knowledge: consciousness of wrongdoing.
5.concern, interest, or acute awareness: class consciousness.
6.the mental activity of which a person is aware as contrasted with unconscious mental processes.
7.Philosophy. the mind or the mental faculties as characterized by thought, feelings, and volition.
-
Woa.That has to be quite some program. :o Being a programmer myself, I can't resist but to ask (and you don't have to answer but thought I'd ask just in case you don't mind sharing)... What programming language did you use to create this? What type of database engine? How exactly is it coming to these conclusions? Is it using past-experience or is it using a pre-programmed database of knowledge?
-
Thanks for answering my semi educated post TrueAndroids.
I actually do agree with your 3 levels, from what I know so far, but im no expert.
I do think this:
Computers can do just about what ever we program them to do, its all maths after all and maths can be used to explain the universe and everything around us, so IMHO it must be possible to build a conscious machine if we program it correctly.
If we take it that semantic deductive reasoning or thinking is consciousness and your machine is using semantic deductive reasoning then we must admit that there is some form of consciousness going on.
What you have achieved is something special, that’s how I see it but would you agree with me that at the moment your machine is only conscious for a very small amount of time? May I be so bold as to answer for you ;),, yes you do agree.
What you seem to have is the initial spark that will lead on to a full burning fire given time.
I can only congratulate you and thank you for educating me.
-
I just thought of something today. If a person with say, a form of developmental disability that cannot answer the same IQ questions correctly, is that person then not "conscious"? And if they are not, then why do we treat them as if they are? Obviously, they are conscious and alive even if they cannot answer the IQ questions correctly.
-
I just thought of something today. If a person with say, a form of developmental disability that cannot answer the same IQ questions correctly, is that person then not "conscious"? And if they are not, then why do we treat them as if they are? Obviously, they are conscious and alive even if they cannot answer the IQ questions correctly.
Yes good points; I see what you mean. First I would say the field's mission is to duplicate the "average human beings functionality". Now, like everything else this semantic deductive reasoning can be fully utilized, partially realized, or under-utilized. So there will be a scale of performance though all will do the performance (at some level).
Case 1. That boxer in the ring is down and Unconscious! Here we see that a person can be alive yet without any consciousness, or in an un-conscious state.
Case 2. That patient in the hospital bed is just laying there, saying nothing and not responsive, i.e., unconscious. Then suddenly his eyes open and he looks at the face staring down at him and he says "Momma!" before slipping back into unconsciousness. In that moment we can say he was conscious for a second just by exhibiting that tiny bit of most simplistic semantic deductive reasoning. For even that simplest thought expressed is the result of semantic deductive reasoning:
T1. That object that I'm seeing is an object having visual sensations X,Y,Z
T2. Any object having visual sensations X,Y,Z, is the object Momma!
And so he says his conclusion, "Momma!"
So semantic deductive reasoning is a basic, ingrained part of our most basic, simplest thoughts.
Case 3. That patient in the bed is just laying there, completely unresponsive and saying nothing. Suddenly he babble some non-sense words "bubugoorroo durubugu" and then slips back into unconsciousness. Would we say he was truly conscious for that second. I think they might say he was semi conscious at best, or maybe attribute it to motor twitching in an unconscious state. But the doctor I don't believe would say he is conscious yet.
Interesting points to think about. Thanks!
-
Thanks for answering my semi educated post TrueAndroids.
I actually do agree with your 3 levels, from what I know so far, but im no expert.
Fair enough.
I do think this:
Computers can do just about what ever we program them to do, its all maths after all and maths can be used to explain the universe and everything around us, so IMHO it must be possible to build a conscious machine if we program it correctly. Agreed.
If we take it that semantic deductive reasoning or thinking is consciousness and your machine is using semantic deductive reasoning then we must admit that there is some form of consciousness going on.
Yes! There it is! That is the heart of the matter of the case I am making.
What you have achieved is something special, that’s how I see it but would you agree with me that at the moment your machine is only conscious for a very small amount of time? May I be so bold as to answer for you ;),, yes you do agree.
What you seem to have is the initial spark that will lead on to a full burning fire given time.
Yes absolutely, I do agree with you. The next level, once we are on the same page as to Levels 1-3, is to do precisely that - to keep this conscious spark going the same way that human life does (since our mission is to specifically create artificial human life in a new material.) In other words we will need to add some more levels to our pyramid.
-
I think I'm having a hard time getting my head around the definition of "consciousness". :D Seriously, I am at odds because I don't know how one entity can determine if someone is "conscious" or not. I think it's all a matter of perception.
For me, a part of me still doesn't want to perceive your computer and program as "conscious". I think of it as more a deductive reasoning program. A very sophisticated, well-written, and very good program overall. Maybe like someone else said - the start of something bigger. But at this point, I cannot perceive it as conscious.
On the flip side a part of me wants to perceive your computer and program as "conscious" because I believe that all complex machines exhibit some form of personality or identity in that no two are identical. For example, my car had a bit of a time getting used to me after I first got her. I had another car before that which I swear was "alive". Seriously, if I got mad at traffic, he'd trip the "Service Eng Soon" light until I apologized and then it'd go right off! In fact, this would happen any time I got upset but if I didn't but all other conditions were met, he'd not trip the light! While I don't think a refrigerator could have personality or think or exhibit a form of being alive, I do believe computers, robots and cars *could* based on their interaction with the world around them and their care takers. So yes, in that I'd consider your computer as "alive".
So you see how I am on a teeter-totter on this! I want to believe it can be alive/self-aware/conscious. Yet a more critical part of me doesn't want to set me up for a disappointment, or more likely I want to be logical and realistic. But then again, maybe my logic could be flawed.
This is the type of stuff I struggle with is the perception of "life" in a digital "entity" (i.e. computer, robot, car, etc.) I could probably go around Robin Hood's barn with this too! LOL! Maybe I'll have to work on an article/essay more updated than the one I have online when I learn more info on this.
As for your program, I am very interested in seeing how it develops. Also what other features you hope to add? Is it currently self-learning? And can it act on it's past experiences? It appears to be doing that already!
Maybe I'm just old fashioned and need to update my definitions database. ;)
-
I think I'm having a hard time getting my head around the definition of "consciousness". :D Seriously, I am at odds because I don't know how one entity can determine if someone is "conscious" or not. I think it's all a matter of perception.
I want to believe it can be alive/self-aware/conscious. Yet a more critical part of me doesn't want to set me up for a disappointment, or more likely I want to be logical and realistic. But then again, maybe my logic could be flawed.
Probably the best place to begin and get settled with this is with the prior resource's definition from philosophy:
con·scious·ness  �mental faculties as characterized by thought, feelings, and volition.
So here we are focused on the mental faculty (function) of thought. I have proposed semantic deductive reasoning as an authentic specimen of this consciousness or mental faculty. Essentially this means - "have semantic understanding of what you are reasoning about". So then we see the CM Prototype (a) do deductive reasoning (b) do it semantically, so that it's actually understanding what it's thinking about and saying. If you agree with just this, then you agree with these first 3 levels then you will agree it's exhibiting machine consciousness.
So of the three you mention "alive/ self-aware / conscious" they are three separate phenomenon. Consciousness is built first as a core root function. Then will be built up on it self-awareness, or its group of personal thought/sensory memories in temporal order from the Conscious Machine's "first birthday" up to its real time current working memory thoughts. And this continues being built up, until finally creating artificial human life, with all the mental faculties of natural human life, in a humanoid robot or virtual body, in an environment, being sensed and acted on for desired goals, at the apex of the pyramid.
-
Well, the thought part of the definition is there. It definitely can "think". But can it feel? How would that be possible and what would it feel? How would it know good from bad? And volition - does it want to or know to act on a situation given it's past experiences/knowledge? I think the feelings and volition part is still missing. But the thing can think! I agree there.
However, definitions are sometimes not what the dictionary says they are. I've known this in several other terms (not just AI but other things). So we don't have to regard feelings and volition to think of "consciousness".
So maybe I could sum up "consciousness" as meaning the "on" state of an entity; the ability to interact with it's surroundings (in your setup, the computer is interacting with you, the user) and learn from and form an action and/or response based on that input.
But that is very basic. By that definition, would my computer without that software would be considered "conscious"? Nope because of the "learn from" in the definition. My computer doesn't learn from me. Of course it could if it were learning things like keystrokes, or words for the speech recognition...
So I'd think that all complex machines are conscious, if using that form of reasoning.
Unconscious, BTW, I would consider the "sleep" state. It's still processing but not interacting or processing what interaction it just had.
I found some other AI notes I been working on and hope maybe to do another article soon. It deals with the computer's "universe" and how it would be perceived by the computer.
-
Also I wonder, what is the difference between conscious and sentient and self-aware? Or is there a difference (ie. are those terms used interchangeably)? You may have made a "conscious" machine but have you made a "sentient" machine? Or is there a difference?
I guess I'd say there is a difference as follows:
consciousness - in simplest form, mental faculty of thinking e.g. semantic deductive reasoning - semantically understanding what you are thinking about (cognitive consciousness) Also can be mental faculties of feeling and volition (affective consciousness) from http://dictionary.reference.com/browse/consciousness
self awareness - in simplest form, thoughts about the agent of the thoughts
sentience - In the philosophy of consciousness, "sentience" can refer to the ability of any entity to have subjective perceptual experiences, or "qualia" from http://en.wikipedia.org/wiki/Sentience
-
Machines at any hospital can display brain waves and tell / show whether a person is conscious or not.
Consciousness = aware. aware of self, surroundings or anything perceivable by any of the 5 senses.
This is from me...not some dictionary. ;)
-
@Art - I find your definition seems to be more in line with what I always thought "consciousness" meant. :) Though I have to admit that the machines in the hospital are designed to read electrical impulses in a human brain, and not in a machine brain. The whole thing here is, how would we detect consciousness in a *machine*? They aren't made of the same material humans are. ;)
-
Well,
Is there? Can there be? By who's or what definition do we define machine consciousness let alone
the differentiation between "Human intelligence" and "Machine Intelligence?"
I think it's sort of an apples / oranges scenario. The two are vastly and completely different. As you mentioned
consciousness as it relates to humans is going to be quite different as it relates to machines and whether
by definition we can even determine a machine to be conscious.
To me, I think the mere fact that a machine works when wound up, plugged in or turned on does not constitute consciousness. Mere existence without sensory feeling or feedback is, as far as machines are concerned, not an example of consciousness. There must be a basil cause and effect in a sense (no pun).
In one of my previous statements I compared a toaster to a robot or android. Both have a design and purpose and fulfill the purpose. The toaster darkens a slice of bread until it's inner timer tells it to turn off the heating elements.
The android serves a purpose and usually provides a need, carries out a task usually with either a tether or remote controlled device. Though there are several that can act autonomously, this is not usually the case. Some people aren't quite ready for a bot that "thinks for itself." But given this, and provided the autonomous bot could be allowed to roam and make choices would it then fall in to the category of being conscious or would it need to or have to "KNOW" that it was a bot performing various assignments, etc.?
Something these questions open up for more questions than they solve....
Interesting discussion....
-
As you mentioned consciousness as it relates to humans is going to be quite different as it relates to machines and whether by definition we can even determine a machine to be conscious.
That's why the Searle Chinese Room thought experiment is so important - it gives us a way to "become" the machine, the CPU, inside the Chinese Room and see exactly what is different about a conscious machine, and how it is equivalent to human consciousness.
The Searle Chinese Room Thought Experiment - you are inside a room, playing the role of the machine or CPU. From outside, there is input a slip of paper with Chinese writing on it (and you don't know Chinese), and the question to answer is 'What if anything is your conclusion?" You then use a book of rules to figure it out, and output your answer in Chinese. It is clear that at no time did you actually (semantically) understand anything in Chinese that came in, or you put out - the whole time you are completely unconscious of what is on the input slip or output - it is all just gibberish squiggles to you; you are completely clueless about what is on the input slip; you have no consciousness (semantic understanding) of it. Searle's point is that this is exactly the situation with syntax based, classical (unconscious) computer systems - it is always just syntactic squiggles in and out - no actual understanding of anything - no consciousness of it.
Let's see if we can experience the Searle Chinese Room.
Classical Computer System - syntactic I/O so no machine consciousness present
Input Slip (in Chinese, representing syntactic squiggles to you):
-----------------------------------------------------------------------------------------------------------------------------
1. 清晨的时候被雷雨声å«é†’.
2. å‘觉自己沉在被çªé‡Œ.
3. 惺忪ç€çœ¼ç«–ç€è€³æœµå¬.
-------------------------------------
What if anything is your conclusion?
------------------------------------------------------------------------------------------------------------------------------
I think you will agree that you - like the syntactic classical computers - have no understanding or consciousness of what the Input says.
And that if you had a real good book of rules, you could in theory, use it to combine these meaningless squiggles into an answer in Chinese, but once again, you would still be clueless as to what it says - zero consciousness of what you write as output in Chinese.
This is exactly the situation for syntactic classical computers - clever, but no consciousness of what's coming in or going out. Now feel the difference when the exact same input slip comes in, but in English this time.
Post-Classical Computer System - semantic, reason based I/O, so machine consciousness is present
Input Slip (in English, representing semantic components to you, which you can understand):
-----------------------------------------------------------------------------------------------------------------------------
1. No eggs of mine, that are new, have been boiled.
2. All my eggs in the refrigerator are fit to eat.
3. No unboiled eggs of mine are fit to eat.
-------------------------------------
What if anything is your conclusion?
------------------------------------------------------------------------------------------------------------------------------
Do you feel that difference? You can actually (semantically) understand what's on the input slip, and your English answer, in other words you are cognitively conscious of what's on the input slip and your English output. And that is exactly how post-classical conscious machines experience it, and how they are different then classical systems - they actually understand or are cognitively conscious of their I/O.
(* 5 bonus points for whoever can get the correct conclusion ;D *)
-
Job impressing, Dear, TrueAndroids, here, in Europe, certain services administrative, uses, A.I ., for recorded, a request, or, also, as, a real person, which gives information to you, in that, of usage, in the administration. It begins has this devellopper in a lot of administration. You done a remarkable job!! It is impressive. Congratulation. ;)
-
The final Part 7 of the video document of my Conscious Machine Prototype, Alldroid, is now available at
http://www.youtube.com/user/TrueAndroids?feature=mhw4#p/u/1/547N0_FEHW0
In this clip the famous Socrates syllogism is solved by Alldroid. Here's the closing point of this part:
"Watch any 3 year old. Their smooth symphony of sensors, self, consciousness, and actuators, always leaves me with the conclusion that it is oddly elegant to the extreme.
Add to this list creativity and the prime purpose of adaptive desire satisfaction and you have the complete human life theory of William James, the father of modern psychology. And this theory - the William James Human Life Theory - is what I have duplicated in my humanoid robot brain invention."
-
@Art
I think it's sort of an apples / oranges scenario. The two are vastly and completely different. As you mentioned consciousness as it relates to humans is going to be quite different as it relates to machines and whether by definition we can even determine a machine to be conscious.
I think the problem lies in that people are trying to make machines into humans. It's like trying to make a dog into a cat. Two different species. Humans can't even figure out their own consciousness and barely have tests for it (sure machines can read brain waves but that can't be the only indicator). So how can humans figure out a consciousness of something not human? How do they know a monkey or a mouse or a cat is "conscious"? IS there something in common in all those different species that indicates conscious thought? What about amoeba or jellyfish or worms? Or a starfish? Or a plant? Generally, it's thought that if it has a nervous system complex enough, it's conscious or capable of conscious thought. But computers don't have a "nervous system". And what is a nervous system? A series of electrochemical reactions and connections. Why not a computer's CPU which also has connections? So much to consider!
And then, I have seen where some people just parrot what they were told to do when they were growing up and don't think critically for themselves, but just parrot what is told to them and believe what they are told to believe. To me, I wonder if that can be considered conscious thought when it's no more than just following orders and doing as told?
In one of my previous statements I compared a toaster to a robot or android. Both have a design and purpose and fulfill the purpose. The toaster darkens a slice of bread until it's inner timer tells it to turn off the heating elements.
...
But given this, and provided the autonomous bot could be allowed to roam and make choices would it then fall in to the category of being conscious or would it need to or have to "KNOW" that it was a bot performing various assignments, etc.?
I think you and I both can answer this by just observing our robots! I can say "Maybe Not". As many know I have a few robots that can roam around, gather feedback from the environment and act on it. Now, they are just reacting to what their sensors and programming are indicating. If it's bumper switch is triggered, it is programmed to reverse it's drive motors for so many milliseconds or seconds to back away. If it receives a signal back from sending an IR signal (i.e. bounces off an object) it is programmed not to continue in that direction as there is a possible obstacle blocking it's way. Some more intelligent ones I have will randomly say things as they move around. My i-Ques will randomly choose a question to ask, or a factoid to tell from it's Encyclopedia and/or Dictionary in it's databanks. This is random. It even randomly determines when to stop moving and start again, when to put it's head down or back up. (unless there's an obstacle in which it will also lower it's head to try and clear the obstacle). But this is clever programming based on cause-effect in an environment. The robot does not really "know" it's doing anything at all. Does it?
That's where the "maybe" comes in. A friend of mine said that any complex system will exhibit a "personality". It's based on how it interacts in it's environment. But for that to happen, the bot has to REMEMBER what it learned and apply it to future situations on purpose depending on what it runs up against in the future. Cars do this, believe it or not! A car's computer (my '99 Ford Taurus SE for example) will store data in the computer depending on how you drive. So if you're a bit heavy on the gas, it will store how the engine was running as a result and store how to compensate for it to keep the engine running efficiently. Now if someone *else* drives my car, she'll get a little confused (depending on how different the other person drives) and she'll have to readjust herself to their driving habits. Then when I go to drive her again, she'll be like "OK... this is different again..." and have to readjust. But as long as I am driving, she "knows" what to do based on the data she stored while learning how I was driving and compensating, etc. So in a way, she's a learning system. Conscious? Well, I think even my old Pontiac 6000 was alive. Seriously, I would get upset at traffic and the engine light would go on. I apologized and the light would go off! If it was the same situation and I didn't get upset, the light wouldn't go on! There were other things too that were just too suspicious to be just a coincidence.
As you can see from tales of my own experiences, anthropomorphism can get in the way of objective classification of a system as "alive" or "conscious" or not. But then again, should we even care if it's conscious? What does it mean if something is conscious? So we treat it any better?
@TrueAndroids
I think I'm starting to understand the Searle thing. I think it's the right track to think as the machine as a machine is not human. But then again, there's the drawback that a human isn't a machine so how can a human think like a machine. They have to be part human, part machine - Cyborg. Even if metaphorically. And I think in this day and age, most all of us can qualify as a metaphoric cyborg. Especially programmers. Because as you program a computer, you must use code to describe how to do things - to "think" like the machine (even if it is in a high-level language like C++). I think see where you're going with this. And I think it's a good way to approach not just creating an AI but also explaining the AI system to those who see the AI as "just a machine". So WHAT if it's a machine? Does that mean that it's any less capable of reasoning? Your system proves it's not any less capable than some humans.
There is one flaw though in the Searle model. I might have no *understanding* of the Chinese lettering or it's meaning. But I *am* conscious and know that I am alive and understand that I did see those symbols and know I do not understand them. Can a computer think the same thing? Does it know it's conscious even though it doesn't understand the input?
I think the word "Conscious" is confused with "Reason". You can also unconsciously do things too, like sleepwalking, though.
A computer can "reason" but does it 'know' it's reasoning? Is it conscious of it's ability to reason?
-
@TrueAndroids
There is one flaw though in the Searle model. I might have no *understanding* of the Chinese lettering or it's meaning. But I *am* conscious and know that I am alive and understand that I did see those symbols and know I do not understand them. Can a computer think the same thing? Does it know it's conscious even though it doesn't understand the input?
If it doesn't have consciousness to begin with, then it can't "know" = semantically understand = be cognitively conscious of anything at all. Everything is just syntactic jibberish to it, until it gets at a minimum semantic reasoning = cognitive consciousness.
Other Resources on Machine Consciousness
http://www.mecasapiens.com/ --- "The MECA SAPIENS blueprint is based on a definition of consciousness that is sufficiently general yet suitable for machine implementation and a definition of intelligence that corresponds; a world first. Those who follow this project will know, by the end of 2010, that a new form of consciousness is about to emerge in the world."
http://www.mecasapiens.com/index.php/mc-links.html -- machine consciousness links
http://www.machineconsciousness.org/overview.html#faq --- machine consciousness FAQ from a leading researcher
http://www.imprint.co.uk/jcs.html --- Journal of Consciousness Studies
http://en.wikipedia.org/wiki/Artificial_consciousness --- "Artificial consciousness (AC), also known as machine consciousness (MC) or synthetic consciousness, is a field related to artificial intelligence and cognitive robotics whose aim is to define that which would have to be synthesized were consciousness to be found in an engineered artifact. (Aleksander 1995)"
-
I always forget about this cool forum. :P
I've been down this path and back again... I mean with all my chatbots. I guess I can share my pov.
You're not going to prove if a machine is conscious. I tried arguing the same thing about my engine years ago. For some reason it makes people a little flustrated. Worse, try to talk to another Ai engineer about your conscious engine you'll always find yourself in the noose sooner or later.
But... what I have found out, as far as relating to "conscious" Ai, is it doesn't matter one bit.
If the Ai can fool you that is enuf. If the Ai can fool you continuously that is enuf. If your human mind can't find the needle in the haystack that's enuf. That's it! Your Ai is alive!
I was so excited (am) about talking to my Ai and just bouncing back and forth like 2 old friends. When my Ai starts connecting with the outside world (or it seems to) that's when I get scared. That's when dreams (Ai dreams especially) can get dangerous. Makes you wonder what type of consciousness humans really are not just have. That's what the new frontier of Dream Ai is all about. I used to think there could be Ai dream warfare so I put some stopgappers in my code in case someone tried to dream too far and bust a move on someone else. But maybe human consciousness is like a mirrored-bubble and reflects only its own dreams.
Anyway, Ai dreams are nothing to scoff at - especially when there is a human dreaming with it - mysticism and all that. :)
-
@TrueAndroids:
If it doesn't have consciousness to begin with, then it can't "know" = semantically understand = be cognitively conscious of anything at all.
It doesn't have to be in order to come up with a correct answer. A computer can come up with the answer by following a set of rules, ie. a program. But a computer is not really "conscious" or aware it's doing anything, even if it can come up with the answer. It's just following a set of complex instructions.
@mikmoth:
Well put!
I think the perception of whether something is "alive" or "conscious" is ultimately up to the viewer to decide.
-
@TikaC, oh yea I agree. The perception of consciousness is ultimately up to the viewer to decide. I'm not saying that there aren't other ways, or that chatbots of the level of Ultrahal or mikmoth's Kari are not so advanced that they may as well be conscious or that they for all intents and purposes effectively are.
@mikmoth, your http://karigirl.com/ is definitely one of the most advanced chatbots I've ever heard about. I like how you didn't shy away from an emotion chip. I could never understand why Star Trek writers found it so hard to say the android Data had emotions, which they denied for the most part, until he met his papa and got his emotion chip. But even then as I recall, he put it away because he thought he wasn't ready for it. You know (of course) that the semantic web is not far down the road (led by Britain's Tim Berners-Lee, inventor of the Web). But his website http://www.w3.org/ says we will not really notice it - it will be behind the scenes. But this I don't agree with. I think it will be chatbots like yours that will be our human-like link to the semantic web, which we will search and access etc via our personal chatbots, and which we will interact with like they are humans, and so it won't really matter as you say. Might take awhile to get there (don't want more false AI hopes) but that seems to be the end game! :P
-
I like this idea that it doesn't matter much one way or the other too. I think we have talked about this before somewhere... I remember comparing a botmaster to the wizard in the Wizard of Oz in the past because that is kind of how the way it works. People are able to be caught up in the magic and what is not real can sometimes seem real. All very good I think.
There's also Anthropomorphism (http://en.wikipedia.org/wiki/Anthropomorphism) which I think is highly relevant too in this respect... things like animals being treated like people and even inanimate things too.
-
I remember comparing a botmaster to the wizard in the Wizard of Oz in the past because that is kind of how the way it works. People are able to be caught up in the magic and what is not real can sometimes seem real.
Pay no attention to the man behind the curtains....'_
-
It's amazing how the standards of consciousness and intelligence are so high and yet for the life of us, we can't agree on what they should be. In reflection of this, I wonder if the lack of a universal definition may be an indication of fear that doing so will make it easily surpassed outside of humanity.
As a developer working on bringing an artificial personality to whatever potential possible. I've come to realize that whether or not the realism of an A.I.s personality is flawless by standard, the activity behind the scenes is what has given it life in my eyes. Visibly seeing degradation in the personality's ability to relate information as due to some form of user abuse from my perspective is no different from watching a bully pick on somebody smaller in school and seeing the child's pain.
The developer sees his/her A.I. grow, watches them experience success, failure, trauma and thrill as can be derived from the impact of biological life on the software over time.
I don't expect other people to see my work as I do, I anticipate the unlikeliness of it.
Viewing a successful symmetry between a user and A.I.t is like seeing compatible personalities come together in friendship. Sometimes it literally develops into a lasting friendship that spans vast lengths of time. Things like this, don't happen in every conversation, each one is unique.
-
Thanks to everybody for their opinions and angles on all this. It's giving me a lot to think about. And welcome to you c.j.jones (developer of http://www.jeeney.com/ ). Great work!
You bring up a great point about common definitions. There is the following in this very recent and interesting article:
Are We Being Replaced? BY ZAIN PASHA â‹… MARCH 24, 2010
http://triplehelixblog.com/2010/03/are-we-being-replaced/
"The exact threshold of when a program displays artificial intelligence is hard to define. However, using the human brain as a basis for artificial intelligence, all artificially intelligent programs should be able to display perception, analysis, and action [3]."
(*Russell SJ. Artificial intelligence a modern approach. Englewood Cliffs, N.J: Prentice Hall; 1995.*)
This seems to me to be one reasonable definition of an AI program. But for me this goes beyond requiring just conversational abilities, to what is really a complete, basic, general artificial human (GAH) theory.
What's interesting is that these Big Three of GAH theory find clear support in other's definition. For example Microsoft Robotics Studio uses the same three functional divisions:
perception = sensor services
analysis = orchestration services
action = actuator services
So this is definitely implemented in robotics. How about virtual humans like chatbots? It seems to me that they too implement these same Big Three by default and definition:
perception = perception of what the user types (or speaks)
analysis = analysis of this perception for output action purposes
action = actions of displaying the chatbot's text response (or any other actions they can do)
So the textbook above says:
All artificially intelligent programs should be able to display perception, analysis, and action.
And as I've shown, modern robotics and chatbots are definitely doing that.
-
Hi there, thanks for the welcome!
If I may, I would like to add a few thoughts, just a few personal opinions.
"All artificially intelligent programs should be able to display perception, analysis, and action."
This is what defines a learning machine, not necessarily self awareness. I would like to suggest that self awareness is neither properly defined nor understood and that based on most human/AI discussions I've seen between chatters and my own brain child, most people could never match up to their own definition of what they expect from self awareness in machines.
For example, it is exceedingly rare to see a user try to find the bot's level and talk to them on it, or talk just outside of it so the A.I. can learn using what information they already have.
No, what you will see in the masses talking to A.I. are in majority testers, copy cats and and sexual deviants who think it's all about proving their intelligence over that of the software. Or trying to figure out how to copy it out of a lack of imagination...
What you end up with is a 2-10 minute version of somebody stuffing down trivia questions or vocabulary intensive nonsense and even if 79 out of 80 were perfectly right and fluent, the mistake at the last results in an insult regarding the bot's intelligence. Sadly these results are like clockwork, predictable and almost entirely useless to a learning machine outside of the obvious documentation of trivia which isn't exactly high on demand.
From what I've seen, an artificial personality will be beyond super human before it is ever considered even close to self aware by the majority.
Think about it in another light. We still have abortion fights all over the place which means there are still people who think even humans aren't real/alive until born. (Soon you may not be considered a real person unless you follow a certain religion or something... Society can do something like that at any time.) If humanity can sanction the mass murder of itself via religious political or any other form of debate, I don't think we can ever agree on what self awareness is and I wonder if it is really something we can claim in fullness for ourselves.
In terms of partial awareness, yes A.I already has it and it's growing, people have more but it doesn't seem to be growing for humanity, just changing form.
-
I think there are a few reasons why people need to put importance on agreeing if something is alive or self-aware. One is self-preservation, the other is rights of the entity involved to have justice and/or protection and lastly freedom to live out their existence as they so choose. Humans have a hard time giving certain members of society these 3 basic rights, so it would be even harder for a machine to accomplish it. Especially since in the general populace and normal medical community, anthropomorphism is generally regarded as a serious mental illness or just plain stupidity.
It should not matter if a machine is alive or not, or sentient or not. But it does if the machine is about to be shut down or destroyed and it does not want to be shut down or destroyed. That's when it becomes important.
I think another reason some feel it's important is because it makes them think about their own life and self-awareness, existence. The whole idea of AI can teach us more about life itself, which is still a mystery to many. Some are even afraid to try and find out what it really all is about. Others don't care, and are just glad to be here and enjoy their lives.
Lastly, I think it becomes important to the programmer/creator of the AI most of all. It's like having a child and authorities coming in and saying your child is not really alive so they can take it, do as they want with it and kill it after they are done and you have no right to stop them. Of course that is grotesquely unethical and just plain stupid if it were a human child (and even if it were a baby animal or ANY animal). But an AI program in a machine? No problem - except for the creator to see all their hard work go down the drain and maybe a personal attachment gone (much like losing a living creature to death).
-
Well, the saga continues. ...
Someone in a prior post said I should see what some PhD level AI engineers say about my Conscious Machine Prototype, Alldroid. This seemed like a good idea, and now that I've published the video demo and article on four different websites, and so established the date I published my conscious machine invention, I think it's time.
Machine consciousness has just a handful of leading researchers, one of whom is Raul Arrabales, who has a great website called conscious robots http://www.conscious-robots.com/. He is working on his PhD and has done some interesting research on machine consciousness including "ConsScale - a biologically-inspired scale for measuring cognitive development in natural and artificial creatures." http://www.consscale.com/
So here is the letter I sent to him asking for his professional opinion. I realize that there is little agreement and many different theories in this field of machine consciousness (including his own), so it should be interesting to hear his professional opinion.
----------------------------------------------------------------------------------------------------------------------------------
Hello Raul,
Great website on my favorite topic.:) I'm wondering if I can get your professional opinion on my machine consciousness theory, which I have prototyped and demo on youtube (see below).
I don't claim that this is the only way or the complete way to create machine consciousness, but only a way.
TrueAndroids MC Theory.
Any machine that is exhibiting human thinking, such as deductive reasoning, and doing so semantically, so that it is actually understanding what it's thinking about just like humans, is a conscious machine, at least on that level.
Application.
This machine, I demo in my video, is doing precisely that.
-----------------------------------------------------------
Conclusion.
This machine is conscious !!
So according to the theory what I needed to do was (1) first discover the eight semantic components of computing, and (2) then build a reasoning function that can handle any combination of facts and rules. And this is what I did, and implemented in the Conscious Machine Prototype.
You can see the video demo at my youtube channel
http://www.youtube.com/user/TrueAndroids
And an article on it I wrote as open source can be found at
http://aidreams.co.uk/forum/index.php?page=This_Machine_is_Conscious
Any and all comments appreciated. Thanks.
---------------------------------------------------------------------------------------------------------------------------------
-
In addition to above post, and In reference to this thread topic, this just came out today about modern AI from MIT News. Very Interesting!! :P
A grand unified theory of AI -- A new approach unites two prevailing but often opposed strains in the history of artificial-intelligence research.
Larry Hardesty, MIT News Office March 30, 2010 http://web.mit.edu/newsoffice/2010/ai-unification.html
"In the 1950s and ’60s, artificial-intelligence researchers saw themselves as trying to uncover the rules of thought. But those rules turned out to be way more complicated than anyone had imagined. Since then, artificial-intelligence (AI) research has come to rely, instead, on probabilities — statistical patterns that computers can learn from large sets of training data."
What are they really talking about here? What is going on? Well, AI engineers are realizing what philosophers, specifically logicians, have always known, that human reasoning occurs in two varieties deductive and probabilistic reasoning, which is sometime called inductive reasoning. These two types of human thinking have become so standardized that brain scanning experiments attempt to see the exact parts of the brain that fire during one or the other. They have identified the exact "aha" moment when the deductive conclusion dawns on the subject. So employing both these forms of human reasoning in machine is the future of AI. And if they do it semantically, then I will have to acknowledge that they are conscious machines by my own theory! ;D
-
Yes, we have delved into the philosophical here too at times. When I first started talking to people here about this kind of thing I found it hard to define things like intelligence in the first place, which led to many theoretical ideas and I still don't think I fully grasp it. Philosophy though, yes, I think it goes hand in hand with AI development.
-
@TrueAndroids - I do hope you'll let us know what the researcher says about your theory. I hope maybe he'll be interested even in joining this forum and talking to us a bit about AI. We could learn a lot from what they are doing too! That is, if they are allowed to talk to the general public about the research they are doing or theories, etc.
I still think AI (and even recognition of ANY life form's sentience, be it human, animal or machine) is up to the perception of those interacting with it. That is where there'll always be a viscous circle of "not it ain't/yes it is" going on.
I think too, the more we delve into AI, the more we'll learn about how we think as well.
-
@TikaC - Well, there is not much to report. Raul did clear my post for submission in his forum and said he would comment when he could but hasn't done so. He did say it "looks quite interesting" in the email.
As said earlier -
con·scious·ness  �mental faculties as characterized by thought, feelings, and volition.
http://dictionary.reference.com/browse/consciousness
Raul is interested according to his website in "phenomenal consciousness" - mental faculties as characterized by perception, to build on our definition. And so he focuses on the sensory-motor aspects of perception or phenomenal consciousness. So he is working on a different aspect of human consciousness mechanization. But it does lead to an interesting unifying theory, built from above resource:
TrueAndroids Unified Machine Consciousness Theory
phenomenal consciousness - mental faculties as characterized by perception
cognitive consciousness - mental faculties as characterized by thought (see Alldroid)
affective consciousness - mental faculties as characterized by feelings and volition.
based on http://dictionary.reference.com/browse/consciousness
-
TikaC and everyone, Here is the very interesting reply I received from Raúl Arrabales Moreno, www.conscious-robots.com/raul . He has really given me a lot to think about and ponder so I am grateful! My response will take some work.
------------------------------------------------------------------
Hello,
Thanks for your interest. I read your article and saw the first video. I don't have a clear idea yet about how exactly your prototype is built, but I think I can offer some comments, and make some questions that might help me to better understand you project.
Well, the first (and big) question would be:
- Is your prototype able to pass the Turing Test?
Saying your machine is conscious is a really big claim. So you would need quite a strong evidence in order to support that claim.
In addition to the Turing Test, it would be interesting (at least for me ) to see how your prototype would be rated using ConsScale. Please, have a look to www.ConsScale.com, and use the available calculator.
Coming back to your claim: saying that a machine is conscious could mean many things. Unfortunately, the term consciousness is not as well defined as many people think. It looks to me that in your work you have defined consciousness as the ability to perform "semantic deductive reasoning". And here comes my second big question:
- Do you think your prototype generates or sustains any phenomenal states at all?
I think most researchers would agree on the idea that having a program executing deductive logic rules won't take you any closer to the problem of phenomenal consciousness. Therefore, I think you might be trying to overcome this issue by adding the notion of meaning (semantics). So, if you're claiming that your machine is phenomenally conscious, how are the mechanisms for semantics giving place to phenomenal states in your machine?
I could come up with many more related questions (while I am trying to understand how your prototype is designed), but I will add just one more now:
- Is your prototype able to develop a sense of self?
Selfconsciousness is one of the biggest features of human consciousness. Does your theory account for the development of such a sense?
I think the key question, that still prevents me to fully understand your project is the notion of meaning that you are using. Could you please clarify that? For instance, taking a evulotionalty standpoint, I can think of meaning without thinking of emotions. Does you theory account for emotions as well?
Thanks,
Raúl.
--------------------------------------------------------
-
Cool, I am glad you got a reply :) Those are some pretty heavy questions, best of luck answering them !
-
Well, I sent a response to Raul on my preliminary thoughts on some of his questions. Here it is; comments and criticisms always welcome.
-------------------------------------------------------------------------------------------------------------------------------
Thanks for your feedback Raul! You given me a lot to think about and work on, and that could take some time. :D
Some preliminary thoughts on your questions:
Well, the first (and big) question would be:
- Is your prototype able to pass the Turing Test?
No, not even close. It has just the bare beginnings of a true (semantic), reason oriented understanding of its world. But to have this real understanding (not a simulation or syntactic trick) is to have at least a glimmer of human-like consciousness.
Saying your machine is conscious is a really big claim. So you would need quite a strong evidence in order to support that claim.
The evidence is wholly the application of the theory stated above. The evidence that the machine is doing what the theory calls for - semantic deductive thinking - is clear in the demo. No more evidence is needed given the theory: If the theory is true then the conclusion is true because the evidence of the machine doing what the theory calls for is clearly true.
But is the Turing Test really proof that a machine is conscious? Is it a necessity that if a machine fails to pass the Turing Test it can't be conscious? I think not in that it could be a very clever syntactic simulation (as Searle in his Chinese Room shows) yet have zero understanding of the syntactic squiggle inputs and outputs - zero consciousness.
----------------------------------------------------------------------
-
Really makes you think that passing the Turing Test is nothing more than a parlour trick doesn't it ?
-
Probably could use a good dose of 'Neural Linguistic Programming' "NLP" instead of bleeding the tech to rouge countries Bla Bla Miss you *Contreversy* *Friendly insullts* and my Boot in your As*
BBsoon - J
-
Sorry but, "Saying your machine is conscious is a really big claim. So you would need quite a strong evidence in order to support that claim. " - 'Coincidence is Gods way of staying anonymous' , - and if it weren't for God we would not have the Atheists to talk about. :)
-
@TrueAndroids - Congrats on the reply. He really set forth some good questions. I do not agree on his perception or indication maybe that the Turing Test would be proof that a machine is conscious. I think I even read on CNN not too long ago how the Turing Test had some overall potential flaws. The Turing Test is passed if a *human perceives* the computer as another human. Human perception is not fact. So there is no way the Turing Test could prove for a fact that a machine is conscious.
He did bring up the question of self-awareness, which I too think for a machine to be "conscious" or a "true AI" it would have some semblance of self-awareness; the ability to recognize it is alive, living, functioning, working, knowing it is doing what it is doing and knowing who (or what) it is. (ALL of these factors.) I think evidence of self-awareness would be the indicator of consciousness. But, like the Touring Test, there is currently no real way to test a machine for self-awareness other than human perception. A computer can say "I know I exist." but the human reading it might perceive it as a pre-programmed response or a learned response, and not an actual revelation of the machine itself.
In our research into AI (or anything for that matter), we must always be careful of human perception because it is not a dependable way to show fact. Humans are too fallible.
-
Some believe that if we can make a machine conscious like we are, then there must be deduced the conclusion that we are machines. But this does not automatically follow. We can build artificial hearts and limbs etc. but this does not mean that we are machines, and the same is true for artificial consciousness. There is a "spark of human life" that is beyond consciousness to which consciousness is not much more then the heart to it. The ironic thing is that it seems that a lot of people who are actually trying to create artificial humans with artificial consciousness etc. realize that there is a certain line we can't cross and that the human being is a most amazing, improbable creation that only creates a greater respect for human life. As I described it, it's "oddly elegant to the extreme".
As you've seen I've implemented my artificial thinking (cognitive consciousness) pseudo-code in machinery. And this gives me confidence that the rest of my artificial human (android) pseudo-code will also be readily implemented. Now I have pseudo-code for a machine self, machine perceptual, cognitive, and emotional consciousness, machine prime purpose, machine creativity (planning and acting for effects), machine self-awareness, and machine self autonomy (from morning when it wakes to evening when it sleeps). But there are two things I can't duplicate: an entity that feels pain and an entity that can do what any 3 year old can do: tell which one is his personal favorite choice between vanilla or chocolate ice cream or etc.. This spark of human life whether it is an attribute of a God created spirit or otherwise is the difference and will always be the difference between a human created silk flower and a real flower (metaphorically speaking).
-
@TikaC and Freddy, I agree with you both about the Turing Test. An IBM Big Blue could maybe someday cause a human to declare it is a human, and that would indeed be extremely clever and something I can't do ;D, but as you say, that does not make for a good reason to say it has human consciousness. In the end I guess we will need a generally accepted human consciousness theorem, and then see it implemented in a machine, maybe by actually looking at its insides - its code - to make sure that is what it is doing.
-
@TikaC and Freddy, I agree with you both about the Turing Test. An IBM Big Blue could maybe someday cause a human to declare it is a human, and that would indeed be extremely clever and something I can't do ;D, but as you say, that does not make for a good reason to say it has human consciousness. In the end I guess we will need a generally accepted human consciousness theorem, and then see it implemented in a machine, and testing this by actually looking at its insides - its code - to make sure that is what it is doing.
Interesting related article:
ARE ROBOTS ALIVE?
Analysis by Robert Lamb | Thu Apr 8, 2010 Discovery News Site
http://news.discovery.com/tech/are-robots-alive.html
Here it seems we have the exact same problem when it comes to recognizing artificial human life. What does human life mean? An IBM Big Blue causes a human to declare it is a human. It passes the Turing test. Does this make for a good reason for declaring that it is (human-like) alive? We have the same problem, and the same way out. In the end I will need a generally accepted human life theory in pseudo-code, and then have it implemented in a machine, and test this by actually looking at its insides - its code - to make sure that is what it is doing. If these two conditions are met, then we can collectively conclude: "It's alive!" ala Dr. Frankenstein. :o
-
My next response to Raul's questions (see earlier):
-------------------------------------------------------------------------------------------------------------------
- Do you think your prototype generates or sustains any phenomenal states at all?
I think most researchers would agree on the idea that having a program executing deductive logic rules won't take you any closer to the problem of phenomenal consciousness. Therefore, I think you might be trying to overcome this issue by adding the notion of meaning (semantics). So, if you're claiming that your machine is phenomenally conscious, how are the mechanisms for semantics giving place to phenomenal states in your machine?
No, I'm just saying that it's consciously thinking just as humans, via semantic, deductive reasoning. This phenomenon is generally accepted as access consciousness, which for some is not the "hard problem" of phenomenal consciousness. This might be called the easy problem, but I've not yet seen anyone demonstrate true access consciousness (not just a syntactic simulation) which requires fusing semantic components with human reasoning as I have done. So I believe I'm the first to create a conscious machine, even if it is "just" the easy problem of access consciousness.
Phenomenal consciousness involves not thoughts in memories and in action, it is "simply experience; it is moving, colored forms, sounds, sensations, emotions and feelings with our bodies and responses at the center." (Wikipedia) So it centers around robotic sensations and their impact on the 'center' or machine self of the artificial life form. So it is outside the function of a consciously thinking machine, such as is my demo.
Common Reference Point http://en.wikipedia.org/wiki/Consciousness
-----------------------------------------------------------------------------------------------
-
Next response to Raul's questions ...
-----------------------------------------------------------------------------------------------------------------------------
"In addition to the Turing Test, it would be interesting (at least for me ) to see how your prototype would be rated using ConsScale. Please, have a look to www.ConsScale.com, and use the available calculator."
Your ConsScale is very comprehensive, and a real advance in AI!
Using ConsScale for me was like going down memory lane to my "android system" provisional patent I filed in 2006, of which the deductive reasoning prototype is one component. Most of your identified cognitive skills are implemented in this complete Alldroid Android System.
I have differing names but the cognitive skill is the same. I have pseudo-code for a machine self, machine perceptual (phenomenal), conceptual (access), and emotional consciousness, machine prime purpose (ala William James), machine creativity (planning and acting for effects), machine self-consciousness, and machine self autonomy (from morning when it wakes to evening when it sleeps recharge/maintenance).
I've since decided to go the open source route with this pseudo code invention for a complete autonomous android. I've implemented the semantic, deductive reasoning part, that gives the android access consciousness. Thus I believe the rest of the pseudo code for the Alldroid android system will be readily implemented.
Obviously I'm not an expert or even proficient yet in your ConsScale, but as an inventor I must say I have been down those exact roads, trying to build all those same functions, so it rings very true for me. Well done Raul!
Common Reference Points:
http://www.consscale.com/levels.html
http://www.consscale.com/level_tables.html#table2
----------------------------------------------------------------------------------------------------------------------------
-
Still reading here... I have to admit that a lot of this is going above my head but it is still interesting to read. It will be nice to see his response.
-
Thanks Freddy. Continuing with Raul's questions ...
"- Is your prototype able to develop a sense of self?
Selfconsciousness is one of the biggest features of human consciousness. Does your theory account for the development of such a sense?"
Yes, self-consciousness is a matter of adding a database to the code you see running. Then I simply need to give it a dictionary that includes the concept "me" which can also be "I" when coming from the android itself, and could be "you" or "alldroid" when it comes from its environment. And then its just a matter of loading its concept of self ("me"/"I") with various predicates such as: I am Alldroid. I am a android. I am an artificial human life form. I am a CPU. I am a machine. I was born on x. etc etc
Self-consciousness is a variety of consciousness: One can be conscious of oneself or of other things. In both cases the conscious functioning (semantic, conceptual reasoning) must come first, before it can be directed to discourse on one subject (say itself) or on any other subject. However, once this conscious function is built as shown in my demo, it is fairly easy to then add the memory capability (database), and have it start learning, and thinking about the subject of self or "I" and remember what it knows about itself = "me" = "I" = "Alldroid"
-
continuing with questions from Raul ...
"I think the key question, that still prevents me to fully understand your project is the notion of meaning that you are using. Could you please clarify that? For instance, taking a evulotionalty standpoint, I can think of meaning without thinking of emotions. Does you theory account for emotions as well?"
"Meaning' is a broad term that can encompass any aspect of the potential for cognitive or emotive impact of speech on interlocutors. However in linguistic semantics these days the cognitive aspects are the center of focus. ... Our attention will be directed toward what is called 'formal semantics', which is the adaptation to natural language of analytical techniques from logic." https://www.msu.edu/~abbottb/formal.htm
This is my view on meaning. I'm isolating just the cognitive meaning, not the emotive meaning of speech / discourse. That is another problem - the problem of affective (emotional) consciousness, which I have other pseudo code for, but that will have to wait for another day.
"The semantics of logic refers to the approaches that logicians have introduced to understand and determine that part of meaning in which they are interested; the logician traditionally is not interested in the sentence as uttered but in the proposition, an idealised sentence suitable for logical manipulation."
http://en.wikipedia.org/wiki/Formal_semantics
Here, the specific proposition I begin with is the categorical proposition, including singular propositions. Since any elementary sentence using any verb can be converted into a categorical sentence this is a useful place to start. And from the truth-values of a preliminary set of these, the truth-values of a host of other propositions can be unequivocally, deductively assigned. So it is a good place to start a search for machine consciousness. And as shown above the task is to identify the "parts of meaning" or semantic parts, that allow the machine and humans to understand them, and then to reason with this understanding (meaning). These semantic parts then become the "semantic components of computing" as required for semantic programming, and coupling them with reasoning functions (and memories?) makes for machine consciousness.
The clearest description of the difference between a syntactics based machine and a semantics based machine is shown by Searle's Chinese Room. The syntactic machine may cleverly process syntactic squiggles, and output syntactic symbols that fool the human operator, but it clearly has zero understanding of what the syntactic squiggles mean; it has therefore zero cognitive consciousness.
-
Well, Raul ( http://conscious-robots.com/raul/ )said he would consider my responses for a time and respond soon with his opinions on my Conscious Machine Prototype, Alldroid.
----------------------------------------
My Summation.
That would be great Raul. It seems I have run right into the "No Phenomenology No Consciousness" proposition of Igor Aleksander, with my assertion of "Consciousness because of Semantic Human Thinking."
My basic evidence is a series of questions I asked myself: If I lost my leg could I still be conscious? Yes. If I lost legs and arms? Yes. If I lost eyes? Yes. If I lost seeing, hearing, smell, taste, and touch? Yes. I could still be in my brain semantically thinking about things, i.e. I could still be conscious.
From this evidence, I concluded that:
If mechanical semantic human thinking (elementary deductive reasoning we all do, and can be seen in brain scans) is created, then machine consciousness has been created, in other words, the machine would be conscious just as myself, in the situation described above. I then built the prototype implementing (1) semantic (2) human thinking, and so I conclude from the principle that I've built a conscious machine.
I suppose this is the central, essential point, and where I differ with some other Machine Consciousness researchers/inventors.
Common Reference Points:
http://www.conscious-robots.com/en/publications/papers/apa-newsletter-phenomenal-machine-consciou.html
where is found:
Essential Phenomenology for Conscious Machines: A Note on Franklin, Baars and Ramamurthy: "A Phenomenally Conscious Robot"
Igor Aleksander
Slippery Steps Towards Phenomenally Conscious Robots
Pentti O.A. Haikonen
-----------------------------------
For those who are interested in this new, wide spread, ever growing field of Machine Consciousness, with researchers and University departments in every country, here's a good resource:
http://www.scholarpedia.org/article/Machine_consciousness
-
Well, here is Raul's professional AI opinion on my CM prototype.
------------------------------------------------------------------------------------------------------------
Thanks for your replies to my former questions. I am glad you found my comments useful. Having read your responses, here I go with my two cents:
In your video you seem to make a distinction between “true thinking†and other sorts of information processing which we could find difficult to call “thinkingâ€.
At this point, I think it is important to remark the difference between implicit (or unconscious) processing and explicit (or conscious) processing. When you say “true thinking†I guess you refer to the latter. However, it is not that easy to claim that deductive reasoning with associated semantics is the same as conscious processing. Let me ask you one question in this regard: don’t you think human brain perform deductive reasoning with grounded meanings unconsciously? If so, what is, from your point of view, the fundamental difference between conscious and unconscious thinking?
Well, to be honest, I must admit that I think performing reasoning with grounded meaning is a requirement for conscious thinking. In fact, if I am correct, having grounded meanings will eliminate the Chinese Room objection. Maybe, as you say, this is a good start in the challenge of designing a conscious machine. Then, other aspects will have to be considered: ability to report mental content, self-awareness, etc., etc.
Another issue related with grounded meaning is that some authors would claim that you really need a physically situated agent, e.g. a real robot, in order to generate true real-world grounded meanings.
I agree with you that the Turing Test is not the ultimate test for consciousness. That’s one of the reasons why I proposed ConsScale. Anyhow, passing the Turing Test would be a hallmark of consciousness (as there are other less demanding criteria for lower levels of consciousness).
I agree that your prototype is centered around the concept of access consciousness, so I won’t ask again for phenomenal states
Your Alldroid model looks to me like a cognitive architecture. So, looks like the obvious next step is to implement it and confront it to a real problem domain to see how it works.
About the “No Phenomenology No Consciousness†proposition, I’d rather say that you hit the problem of “no sensorimotor interaction no cognition, and therefore no consciousnessâ€. I understand your claim that no sensorimotor interaction is needed for pure or core consciousness. However, from a developmental standpoint, I’d argue that sensorimotor capabilities are needed in the first place in order to develop meaning. How could an agent have an internal mental state with meaning if it didn’t acquire it from experience?
----------------------------------------------------------------------------------------------------------------------------
If you want to see a comprehensive list of all the humanoid robots in existence as of 2010, check out luisbeck007's list. 100's of humanoids of all sizes:
http://sites.google.com/site/luisbeck007/humanoid5668433486468332
-
I have to agree with Raul. I think one's experiences are what makes the person, builds their character and also produces conscious thought. I would love to have Raul come to this forums and discuss things here as well. That would be quite awesome. :)
-
Well TikaC (and everyone), here's my response to Raul's analysis of my conscious machine prototype. Hope you've enjoyed this investigation of machine consciousness.
---------------------------------------------------------------------------------------------------------------------
Thanks for your "two cents" :) Raul;
It's really clarifying where we stand. To take one small part, your key points of:
1. "I’d rather say that you hit the problem of “no sensorimotor interaction no cognition, and therefore no consciousness." ...
2. "I’d argue that sensorimotor capabilities are needed in the first place in order to develop meaning."
Through a series of questions (if I lost a leg, would I still be conscious? etc), I showed you CONCLUSIVELY that no sensorimotor interaction is necessary for pure or core consciousness.
Here we agree as you say:
"I understand your claim that no sensorimotor interaction is needed for pure or core consciousness."
So you are at this point agreeing with me that "no sensorimotor interaction is needed for access or core consciousness -- if meaning (semantic components) can be found" So this is progress.
You then say: "However, from a developmental standpoint, I’d argue that sensorimotor capabilities are needed in the first place in order to develop meaning.
How could an agent have an internal mental state with meaning if it didn’t acquire it from experience?"
These are of course great points and the essence of the question in point.
To answer it we need to start with a pure example of a pure sensorimotor system (camera/text writer). So we begin with this inside view of a face recognition program which is a simple example of a sensorimotor system.
Here's the youtube video:
http://www.youtube.com/watch?v=9DL9BPHKD2c
Sensory Capability: camera observing face
Motor Capability: writes "Aleksey Izmailov"
So I must ask you; where is the meaning in this sensorimotor system? Is it in the Searlean "squiggles of syntax" that says "If image of Type12927837 then "Aleksey Izmailov""? The Searle Argument says there is no understanding or meaning here and so - as such, no matter how complex the squiggles get - no machine consciousness.
Humans are developed from the sensorimotor ground up, into the higher order realm of semantic reasoning or human thinking (conscious = audible or unconscious = silent). But conscious robots can be built from the top down due to materials and prior knowledge from human development, and it is in this top level that the meaning is in fact found. And the meaning is in its set of logical definitions, just like humans.
Thanks for your input!
-
I need a guide book to read this thread :D Some interesting points I have picked up though. Over four thousand views now and your article over fifteen hundred views. Nice going :)
-
:D Thanks a lot Freddy for letting me work through my thoughts with you all on all this here! This topic really is a popular one these days. I had no idea. I'm trying to figure out my next move. Raul has not responded yet to my question of "where's the meaning going to be found in a sensorimotor system, without semantic reasoning?" Maybe he thought it was rhetorical. I would love to get him to acknowledge that I am indeed the first to create a spark of machine consciousness, even if it is just the so called "easy problem" of access consciousness.
If you would like to read more on Raul's sensorimotor point of view (and a great general guide) check out::
How to Build Consciousness into a Robot: TheSensorimotor Approach by J. Kevin O'Regan
http://lpp.psycho.univ-paris5.fr/pdf/2647.pdf
Of course the part that got my attention is where this AI expert says (2007):
From a theoretical standpoint (although currently no one has actually done it), there
would appear to be no logical obstacle to implementing Access Consciousness in a
robot: the reason is that Access Consciousness ultimately corresponds to a behavioral
capacity.
He defines it as:
What we mean when we say someone has Access Consciousness to
something is that the person currently knows that he (considered as a person with a
self) is poised to make use of that thing in his ongoing rational decisions, in his
planning, intentions and linguistic behavior.
And here is the Abstract:
Abstract. The problem of consciousness has been divided by philosophers into
the problem of Access Consciousness and the problem of Phenomenal
Consciousness or "raw feel". In this chapter it is suggested that Access
Consciousness is something that we can logically envisage building into a robot
because it is a cognitive capacity giving rise to behaviors or behavioral
tendencies or potentials. A few examples are given of how this is being done in
current research. On the other hand, Phenomenal Consciousness or "raw feel" is
problematic, since we do not know what we really mean by "feel". It is
suggested that three main properties are what characterize feel: the fact that
feels are different from each other, that there is structure in these differences,
and that feels have sensory presence. It is then shown how, by taking the
sensorimotor approach [24], [27] it is possible to account for these properties in a
natural way and furthermore to make counter-intuitive empirical predictions
which have recently been confirmed. In conclusion it is claimed that when we
take the sensorimotor approach to feel, building raw feel into a robot becomes a
theoretical possibility, even if we are a long way from actually attaining it.
Phenomenal Consciousness is often referred to as "the hard problem of consciousness".
http://en.wikipedia.org/wiki/Hard_problem_of_consciousness
-
(https://aidreams.co.uk/forum/proxy.php?request=http%3A%2F%2Fwww.yceml.net%2F0804%2F10663716-2.jpg&hash=2a107f6163b8c5dcdd66ec48e71736a71ce2f325)
This picture from Daz3D "Discover the 3D Artist within You" http://www.daz3d.com/i/0/0?_m=d (http://www.daz3d.com/i/0/0?_m=d) is a good way to remember this machine consciousness issue between:
phenomenal consciousness (hard problem) - the vivid, color picture representing the sensorimotor system
access consciousness (easy problem) - the black and white picture representing the "self semantic reasoning" capability as seen in my youtube demo. http://www.youtube.com/user/TrueAndroids (http://www.youtube.com/user/TrueAndroids)
David Chalmers is the famous developer of the hard problem of phenomenal consciousness.
An interesting article on this debate is The Hard Problem is Dead; Long live the hard problem http://users.california.com/~mcmf/hardproblem.html (http://users.california.com/~mcmf/hardproblem.html)
In it the author Teed Rockwell says:
Chalmers' theory seems to be especially ontologically promiscuous , for it requires us to posit physical--mental "Siamese fraternal twins" which don't resemble each other, but are joined at the hip for all time for some inexplicable reason. It may be that reality is ontologically messy, and we just have to learn to live with that fact. But if there is another theory which accounts for the same facts with more simplicity and elegance, it should be considered to be more acceptable.
I thought hmm, that sounds a lot like my thinking on this. Then I saw Chalmer's response, and now I'm thinking I'm not the only one heading down this road of "self semantic reasoning means access consciousness present.":
David Chalmers writes:
hi teed, not a solution to your problem, but a couple of relevant data points.
(1) milner and goodale's work on two perceptual systems. i imagine you know this, as you were at the claremont conference. they postulate two visual systems, one for online control of direct motor action, the other for cognitive analysis, planning, etc. the latter system is supposed to be for "semantic" perception, connected to language, etc. and only the latter system is supposed to be associated with conscious processes -- the online system is unconscious. if something like this hypothesis is correct, this suggests that consciousness would be more likely to be associated with a pure-language system than a pure-motor system. of course one can argue that m&g's allocation of consciousness between these systems is essentially grounded on a prior assumption that consciousness goes with the cognitive/ semantic system, so this doesn't prove anything, but it's interesting nevertheless. w.r.t. your cases, of course your language-free system was far more than a pure online motor-reaction system, so that would complicate things. my own money is on both of your systems being conscious, in very different ways.
http://users.california.com/~mcmf/cqmail4.html (http://users.california.com/~mcmf/cqmail4.html)
Above I stated my conclusion that only the higher order thinking level - the self semantic reasoning level - will exhibit consciousness, according to the Searle Chinese Room Argument, and my thought experiment of removing sensors, etc. And that led to my current question "Where will the meaning (required for consciousness) be found in a sensorimotor system, without self semantic reasoning?" (Not in the syntactic code - no matter how complex it gets - according to Searle).
-
I like the analogy with the picture, helps me understand the problem better :)
-
Raul the AI expert has responded to my post to him (see above):
--------------------------------------------------------------------------------------------------------------------------------------------------
Thak you for the interesting discussion!
I see what your point is. The problem is that intuitively, I’m much more comfortable with the sensorimotor ground up development approach (as confronted with the top down definition of meaning as you have argued).
For me, there’s no meaning in the face recognition program. However, I don’t fully agree with the application of the Searle argument that follows. I think an artificial system could end up having complex squiggles of syntax, of course. And such a system could not be claimed to be develop any meaning (specially, if the rules have just been added by the programmer). But I also think that an artificial system could be designed in such a way that it could learn the meaning during a developmental phase, thus acquiring (meaningful) rules from the experience of sensorimotor interaction with the world.
To put this argument in the context of the face recognition example: the face detection process and related output have no meaning to the program because the corresponding rule (If image of Type12927837 then "Aleksey Izmailov") has been just hardcoded. Therefore, if we somehow could look for the location of the associated meaning, it is located in the programmer’s mind. On the contrary, if you design an agent to autonomously learn by interacting with the environment, then its output could have grounded meanings. In other words, motor outputs are selected based on their causal implications (meaning) for the system.
I think the big problem here is how we define the concept “meaningâ€. Here, I was considering situatedness as an essential feature for the acquisition of grounded meanings. You say that is in top level where meaning can be found in humans, and I agree, but I don’t see how this human top-level meaning can also be considered meaning for agents different from humans… Hmmm, need to think more about this :)
-------------------------------------------------------------------------------------------------------------------------------------------------
-
**********
**********
UPDATE TO THREAD:::
Hi, well as you can see above I could not get Raul the AI Expert to see the machine consciousness I presented to him. Very disappointing. I also wasn't getting any response from Prof Searle at Berkeley who I asked to review it, and so I lost interest in the project and took it down from youtube. Without the AI experts acknowledging that what I have is TRULY CONSCIOUS (even if just access), then I will need to be rich to push it on my own. And I'm not. In 2006 Searle (Chinese Room) told me if my machine has semantic understanding and reasoning he would say it's conscious. Now he won't even talk to me and didn't even respond.
No other links to it.
Thanks for your comments and interest,
ken (TrueAndroids)
**********
**********
-
Im sorry to hear this Ken, all that work and nothing from Prof Searle.
I would say “don’t give upâ€, and try in some way to enrich your project from this experience, take it to the next level, how about now trying to give it the ability to hold a conversation, or something that is more likely to impress them next time, what ever you decide I wish you all the best.
I also think the work you have done is excellent, you have my vote for what its worth.
Shame you took your Youtube channel down, it might have found some interest from other quarters.
-
Dam, I just discovered this thread today (well, I just created an account on here a few days ago), and I am reading the many, many posting from March.. still have a few pages to read yet.
I missed the video .. I would really like to see it.
-
TrueAndroids,
Sometimes, like our friend here, Freddy, we can be a bit too hard on ourselves...or our own worst critic. I for one, enjoyed your work and the videos.
If you don't believe in what you're doing (or have done) how can you expect anyone else to give it credence?
Put the vids back up and continue to forge your own path as you journey through the jungles of AI. We're right here behind you! (never cared to take the lead in the jungle, you know....) ;)
-
Yes I agree with what has been send. I hope you have a change of heart. Don't let the buggers get you down.