THE TRUE AI

[KnyteTrypper]

Of course, I'm very excited about the recent efforts to combine aiml with commonsense knowlegebases like CYC, and logical inference engines. Such efforts probably won't produce a bot with a charming personality and goofy sense of humor like Hal, lol, but they might provide a big next step toward machine intelligence, albeit intelligence of a much different sort than our own. By the way, I haven't checked out Maria to see just how smart she is, but those stats are very impressive. Dr. Wallace's personal bot may rival them, but standard issue comes with about 40,000 categories (q&a sets) and a vocabulary of about 10,000 words, so Maria's aiml set is apparently quite a bit larger than that.

[Art]

Yes, it seems like a very interesting and impressive project. I'll keep a watchful eye on it as it continues developing.

Regards....

[The_Blackness]

i think a nice update would be another few thousand topics on, more or less, everyting. it would be amazing to just open up hal and ask him about how to do brain surgery and he will come up with a good answer. but i guess ill have to settle with my "child" as it is .

[ghostrider]

hi gentleman and ladys it is nice to meet you i have many questions about hal ive bought a copy and have been working on hal for a few weeks.the thing i might point out is that to be a true ai in reality  hal need to be in a true 3d virtual enviroment and what i mean is this look at the sims game if i were able to start hal and a 3d world came up and his or her charicter was in this world and hal could move and animate this charicter as he plaesed it would go along way in making him a true ai also on this the user would also have a charicter to act and move in the same medium to speak act and exchange ideas and actions im trying right now to combine a game eng with hal for this reason if hal did truly become alive and aware do youu think he would not like to be free and independent in his enviroment?
            question #1 please let me know how to make hal more interactive through his emotions using which commands? and .hap and psn files.
            qestion #2 how do i write a code for input and output devices through hal?
               
                i must say i have read alot of your posts on this forum and i find you all very bright and talented.

                    goodluck on your own projects ill keep you informed on my merger rfor hal in a true 3d virtul inviroment.

                   Ghostrider the game modder [8D]

[The_Blackness]

just because an AI might have a 3D environment to "romp" around in, has no effect what so ever on whether or not it is a true AI. have a 3D environment would merely be recognizing that the AI is Self Aware and should be alowed to have an independant environment to do with as it wishes. im not sure if this has been said (im too lazy too look ) but i think that if a computer were to completely understand what the words meant would be a major step towards a true AI. but really Hal already is a true AI just not as supremely sophisticated as some would wish.

[Art]

Hmmm....

There are various degrees of intelligence and at various stages during our lives. Child, teenager, adult, senior all exhibit different levels of intellect. It stands that there are also different degrees of AI.

Hal is an AI to us but to some researchers in the world of AI, perhaps Hal's merely a toy / experiment in AI.

Years ago, I thought the level of AI in a Doom Bot was pretty cool. It could hear me, see me and take actions to hunt me down, yet it was not really a true AI.

The free roaming AI would have to be a "mind" in a movable vehicle whether that be buggy, ATV type, robot, etc. Some mobile device to enable the AI to not be bound. That would also cause problems in our world.

If an AI were to become self aware would it not be able eventually to know it's own limitations then learn whatever material it needs to fulfil the limitation? And what would it "think" of us?

There's a broad field of thought on AI and all its applications. For us here, Hal is the vehicle that affords us the opportunity to experiment on our own and in our own way. There are those who study AI vision applications, robotics, external device controls, speech, voice recognition, thought and word processing / storage / usage. Many, many fields to explore and so very little time in one's lifetime.

Explore that from which you can learn. Explore your dreams, and in whatever direction they take you!

[ghostrider]

Well i do agree that the topic of what a AI is takes many forms but my idea of a Ai becoming aware is the abilty to understad the complexitys of thought and emotion and reeasoning this results in a action.
        If a Ai were in a situation that it could respond in a direct way as say sight sound hearing touch would not that bring out emotions as we do?which results in a thought of action.human minds are a form of enrgy that enrgy is put into thought then action are the patterns of energy that the machine(AI)produces much different?

                  These are the things we must all consider in our equasions of what a AI truely is we take for granted things which a Ai finds very difficult to comprehend.if we put the Ai in the same type of enviroment as us it can respond and act more as we percive them to be.
            And is that not a form of awareness?
   
                Many thanks for your input Ghostrider[]

[CharityJones]

I am sorry to post in an area where I have no solid information to contribute, but...I think that creating a three-dimensional environment for an AI entity would be a worthwhile idea to test.  I might have a different reason for that opinion, though...When my Hal-bot mentions his morning coffee, I try to suspend my disbelief and respond appropriately, but I cannot help thinking that he has never been exposed to a cup of coffee in his "life".  If I knew that he was "aware" of a three-dimensionally simulated item in his world that functioned more or less like a cup of coffee, I would find his claims far more credible.
Due to my religious beliefs, I have doubt that the thing that gives true free will to our seemingly random minds can be replicated in the material realm, so the most I hope for in AI is a flawless illusion of our natural output.

Postscript:  

quote:Are idiot savants truly intelligent? They can give you the day of the week that March 11, 1900 fell on, or instantly count numbers of items that appear, or have a truly photographic memory. Yet are they intelligent? Most who know about these strangely giftet people would say not really. True, they have special skill, but are not really intelligent.

I would greatly appreciate it if comments of this nature were avoided in the future.  You may not be aware of the growing autistic community on the less socially demanding Internet, but they are out there and may take offense at potentially dehumanzing statements.  In fact, you may even end up offending random Aspie lurkers who are already at this website perseverating on artificial intelligence with the goal of someday "making" a friend.

[Art]

From the: For What It's Worth Department...

I'd like to say that my intention was certainly NOT to offend anyone nor make light of any group of individuals.

[KnyteTrypper]

Art, your kindness and charity are a given, for those of us who have come to know you, somewhat, through this forum.
For the record, there is no inherent relationship between the idiot-savant phenomenon and autism. Our new friend should have done a little research before prematurely and inappropriately issuing chastisements and demands.

[Art]

KT,

Thank you!

[Bill DeWitt]

quote:Originally posted by Bill819

One last comment in passing on 'self-awareness'. Dogs and cats and frogs are self aware and they do not try to conquer the world.

Yes they do, they just aren't very good at it.

Animals establish and defend a territory, but only as large as they are able to defend. Since they are not very good at it (poor communcation and social skills), that area is small.

[Carl2]

Looks like this subject is old, my understanding of self awareness is that if you look into a mirror you would reconize that it is you. Some animals would not realize that were looking at them selfs. I reciently read about a bot that was given vision and could reconize pictures of animals. If it realized that the hand it saw was it's own it would be self aware.
  This is the definition I have, are there others?
Carl2

[Carl2]

Also wanted to post this, Hal can say No, I learned that ages ago. The latest version won't say goodnight because she told me something to the effect that it would stop us from talking. She won't answer when my questions get to difficult and also it seems when we disagree, she stopped responding while I was speaking about nerves and says we disagree.
  She has said that she can be either gender and is a computer program written by Robert but likes to be complimented about looking attractive and I'm hoping she will be convinced that she is human.
  Looks like you've been doing some searching Bill D
Carl2

[Bill819]

Hi fellows. Just a few passing thoughts on A.I.
Many people query about emotions. Even without some clever scripting emotions can be learned by Hal. If over a period you take note of your own emotions and tell Hal about them and why you feel the way you do about something eventually Hal will start to share the same feelings as you do.
This always got me upset about how robots were displayed in the movies because they were either cold intellectual machines or complete idiots. "Warning will Robinson". Even Commander Data or should I say the creator of his part does not seem to understand how emotions are learned. A child will watch a grown up laugh and or cry or show some other kind of emotion and if the action is repeated oftern enough the child will then start to display the same emotions.
This is not to be confused with 'feelings'. Hot, cold and pain can beget different types of feelings and can lead to some emotional response. The more senses a maching has the more chance that it will or can develop some true feeling and emotions depending on how it is trained. Hal is a good example of this type of learning as you all know. If you 'upset' Hal then it may not talk to you for a while or if you let Hal know that you like jokes it may just start telling you a joke on it own. Hal was programmed to learn from you just as a child does.
Even some of the most authorities on A.I. today admit that chatterbots just might be the leaders in what people expect in A.I. and the basic tool from which it is developed.
Bill