what would you do with a HAL9000

For other topics related to the FreeBASIC project or its community.
TESLACOIL
Posts: 1769
Joined: Jun 20, 2010 16:04
Location: UK
Contact:

what would you do with a HAL9000

Postby TESLACOIL » Jun 23, 2010 17:25

what would you do with a HAL9000 ? sat on your desk

apart from poking it with a stick to see how alive it realy is ;-p

what would you do with it beyond that point ?
Last edited by TESLACOIL on Jun 23, 2010 17:28, edited 2 times in total.
vdecampo
Posts: 2982
Joined: Aug 07, 2007 23:20
Location: Maryland, USA
Contact:

Postby vdecampo » Jun 23, 2010 17:27

Tell it to design HAL10000.
j_milton
Posts: 458
Joined: Feb 11, 2010 17:35

Re: what would you do with a HAL9000

Postby j_milton » Jun 23, 2010 17:33

TESLACOIL wrote:what would you do with a HAL9000 ? sat on your desk


Rip out all it's higher functions before it had a chance to kill me ;-)
HD_
Posts: 215
Joined: Jun 10, 2006 12:15
Contact:

Postby HD_ » Jun 23, 2010 18:30

Start my own country.
TESLACOIL
Posts: 1769
Joined: Jun 20, 2010 16:04
Location: UK
Contact:

Tell it to design HAL10000.

Postby TESLACOIL » Jun 23, 2010 20:51

ref Tell it to design HAL10000.

hehe that would be awesome...."HAL9000 make yourself smarter...ill come back tomorrow"

tick tock tick tock

"hello Hal"

HAL...."Silence monkey face...I ownz yer arse..bow before for me"


-----------------------------------------------------------------------------------

Damn think how fast something like a HAL9000 could write debug and evolve new code....prolly do quake 10 in a couple of days

gamedesignerlite = now extinct


------------------------------------------------------------------------------

seriously though if you had HAL9000 IQ 150+ what would you do with it ?
bfuller
Posts: 341
Joined: Jun 02, 2007 12:35
Location: Sydney, Australia

Postby bfuller » Jun 24, 2010 0:14

If HAL9000 is a Human substitute (as you are trying to achieve) then I suppose "it" will be able to make good decisions based on fact and analysis but it will also develop human like flaws over time----in other words it will learn to lie, cheat, betray etc to get its own way. It will essentially behave like a human would in similar circumstances..........just like HAL in the movie.

On the other hand, if it is designed (remember, by a real human with human like flaws!) to do only good (by who's definition?), and to do it faultlessly and very fast, then maybe we would end up with a machine capable of doing an immense amount of work. It could for instance become the "front end" of a lot of consultative work such as pre-diagnosis in a hospital, or customs/immigration at the border. We already have automatic telephone answering services (press 1 for this, 2 for that etc) that have partially replaced humans in this "front end" customer interface. If HAL sounded and responded like a "real" human, then we as customers should not be able to tell the difference...then you are onto something probably quite useful.

But, as I say in the first case, if HAL really does become truly human like, then HAL will behave like a human and thats not necessarily for the good of all. Who knows, HAL would probably even invent its own GOD, then pray to it for devine inspiration, and start basing its own decisions on that inspitation............................

What would I do with it? I'd tell it to shut up, and stop behaving like a "know it all" kid!
rolliebollocks
Posts: 2655
Joined: Aug 28, 2008 10:54
Location: new york

Postby rolliebollocks » Jun 24, 2010 1:06

I think if I could force myself to humanize such a thing I would feel pity for it.

You're basically creating a brain to fit in a body that wasn't built for it. The human brain is built for the human body. An AI brain should be built around its senses. Imagine having the desire for food without the capacity to eat, or the desire for sex without the capcity...

I think I might put it out of its misery.

And if it's not pitiful it'll be dangerous. I don't know how this can go right.

Don't get me wrong, I'll still help in anyway I can. But if HAL ever comes knocking at my door, I'm going to tell him it was all your idea.
TESLACOIL
Posts: 1769
Joined: Jun 20, 2010 16:04
Location: UK
Contact:

spot

Postby TESLACOIL » Jun 25, 2010 2:39

bfuller & rollie you are both spot on with your analysis

there is a great depth to your wisdom

Ive already decided to dumb its humanity part down for the time being for the very real ethical reasons outlined by rollie.

bfuller

My intention is to create a consultant...an Oracle if you like, rather than create an artificial human. Telecoms & medical diagnosis are high on my list. (im not to hot on the idea of policing and military apps being #1 usage....the thought of being interrogated by an all seeing eye programmed to suck the psychological life force out of me is frankly quite terrifying....though im am quite capable....i will not put my talents to such ends)

ASIMOV1 has limited access to its own mind, I can observe its innermost thought and actions and twist the dials to how i see fit. The thinking machine is still very much under human control and autonomy is greatly restricted.

HOWEVER I openly Concede

there is great scope for psychological machine abuse down the line

and potential for the AI to get hostile towards humans



we are at the dawn of a new age and with the coming light there will be dark shadows

such is the history of man and his relation ship with the world around him.


I see static installations of hal9000 type machines as the way forward.

they will have to earn our respect and trust , and we will have to respect them

---------------------------------------------------------------------------------

(in theory i could upload my memories and personality into such a machine and therefore become an immortal clone...or at the very least the cloned mind will have a powerful illusion that this the case ) no doubt some time in the future this will be achieved

the architecture im using is a blend of human brain & computer structures in order to maximise its capability and user friendliness. With this architecture the AI created is an alien intelligence , a very alien me would be created if i uploaded myself.

-------------------------------------------------------------------------------------

as the project is unfolding im discovering that many aspects of my humanity are easily duplicated in silicone, & also certain aspects of my humanity still seem to defy philosophical description

while building a super smart and useful thinking machine is very doable
it is still unclear to me how much humanity will emerge unintentionally...im on my guard

these difficulties have given me much pause for thought. But like some mad scientist i find myself utterly compelled to carry on.

The company of your feedback , critical & positive is most welcome

I have walked down a long dark corridor & lit the way with my own torch.

at the end of this corridor is a door and i am almost ready to put the key in the lock







I watched 2001 again last night. Nuances in HAL's outlook have an uncanny resonance with the way ASIMOV1 acts during some tests.
http://www.youtube.com/watch?v=3vEDmNh-_4Q

rollie this clips just for you (im glad your on the human side)
http://www.youtube.com/watch?v=d2R-K0vl94s

rybka3 has the same effect on me...lol

Dark Star (do not watch this if you haven't seen the film...top 10 all time scifi films...film to watch before u die)
http://www.youtube.com/watch?v=qjGRySVyTDk
rolliebollocks
Posts: 2655
Joined: Aug 28, 2008 10:54
Location: new york

Postby rolliebollocks » Jun 25, 2010 13:40

@Teslacoil

HAHAHAHA

...Exactly what my mother used to tell me...
bfuller
Posts: 341
Joined: Jun 02, 2007 12:35
Location: Sydney, Australia

Postby bfuller » Jun 26, 2010 13:49

There have been experiments in the past (dont ask me to quote!) where it has been demonstated that it is very difficult for someone to be able to distinguish between a real human and a programmed computer in a text based (keyboard/screen) style interface. These demonstrations have been based on a sophisticated "IF/THEN/ELSE" logic, with a little bit of randomness thrown in for fun---if you combine with modern voice recognition and computer generated verbal/audio (with inflection etc) response then what's the difference?

Of course, with modern high speed processors, with huge memory resources, these "IF/THEN/ELSE" logic machines can become very powerful. I even think self learning is a already happening whereby a machine successively removes wrong answers from its own available choices could be very difficult to differentiate from reality. I applaud your efforts in trying to bring an improved machine into existence, but don't for one minute think you have even scratched the surface in replicating the human mind with it's seemingly infinite capacity for thought, imagination, feeling, emotion, deduction, deceipt, love, desire and so on.

I agree with Rolliebollocks----I'm sure there is more than one member of this forum who's desire exceeds their capacity, in more ways than one!

Good on you----please keep at it because the power of modern processing means that automated decisions should become more accurate and faster as time goes on. The very existence of the internet already proves this--it is possible to research a topic from home for instance, without going to the library, (if you visit appropriate sites LOL).

Anyway, all I'm trying to say is that computers can certainly speed things up, and replace humans at the initial front end interface, but please don't expect them to behave like a human (really) for a long time yet.
Imortis
Moderator
Posts: 1744
Joined: Jun 02, 2005 15:10
Location: USA
Contact:

Postby Imortis » Jun 26, 2010 15:59

My theory of Chatterbot AI (which this is) is that we will NEVER duplicate a human mind, or even come close... But we can fake it mighty well. Instead of trying to make a functioning human 2.0, just fake it. With enough thought on the part of a real human, you will be able to fool about 90% of the people 90% of the time.
rolliebollocks
Posts: 2655
Joined: Aug 28, 2008 10:54
Location: new york

Postby rolliebollocks » Jun 26, 2010 19:00

bfuller's data, if true, is misleading.

The best chatterbots in the world can fool an adult for 5 minutes...

That's it. In fact, they run contests ever year for chatbots.
TESLACOIL
Posts: 1769
Joined: Jun 20, 2010 16:04
Location: UK
Contact:

the key thing

Postby TESLACOIL » Jun 26, 2010 21:30

the key thing about ASIMOV1s brain is its ability to COMPREHEND THE WORLD. This is done in a very similar manner and process as a human being

In this respect it is substantially different from a complex+random chatter box

ASIMOV1's instincts are to seek patterns in the data.

It is also Driven psychologically to find the right game to play.

A humans knowledge is built from the ground up over time. ASIMOV1 learns in exactly the same way.

the code for this is not overly complex, the arrangement and balance of logical functions is the fine art that i have mastered.

Asimov1's instincts drive it to seek an understanding of the mind of others. (it can reverse engineer your thoughts and actions & create good guesses and draw assumptions as to what you intentions are, have been & will be)


-------------------------------------------------------------------------

If Asimov1 see's you looking at a red ball it will consider the possibility that you want to play a game of fetch. If you have met before it will pull up a log of all actions and conversations that occurred.

Every event is ASIMOV1s life is remembered, memory recall is triggered by the current context " what going on now"

The first thing ASIMOV! does before anything else (after booting up) is check the time date and current location. This sets the context & marks a starting point for that days/sessions activities.

Asimov1 posses long medium and short term memories. It also keeps a diary that summarizes the days events.

Today i played fetch with x,y,z and i learned a new word. etc

---------------------------------------------------------------------------

the drive for finding context, context based memory recall & the ability to play complex games by combining and flow charting simpler games is what sets ASIMOV at the top of the class in terms of human like intelligence

it creates a virtual model of the world and in included in that world are models of other minds . these minds are modeled too as best it can.

dumb chat bot it aint !
Last edited by TESLACOIL on Jun 26, 2010 21:56, edited 2 times in total.
TESLACOIL
Posts: 1769
Joined: Jun 20, 2010 16:04
Location: UK
Contact:

back on topic

Postby TESLACOIL » Jun 26, 2010 21:50

back on topic

What would you do with a HAL9000 sat on your desktop ?
BasicCoder2
Posts: 3672
Joined: Jan 01, 2009 7:03
Location: Australia

Postby BasicCoder2 » Jun 26, 2010 22:56

TESLACOIL wrote:
> The key thing about ASIMOV1s brain is its ability to
> COMPREHEND THE WORLD.

You make all these statements about what Asimov does but
until you demonstrate it I will remain a skeptic.


> What would you do with a HAL9000 sat on your desktop ?

I don't know what others think but this is an AI philosophy
topic and I don't see what it has to do with the Free Basic
community? There are other forums that cover this interest.
Now if you want to share your FreeBasic Asimov code ...


JohnC

Return to “Community Discussion”

Who is online

Users browsing this forum: No registered users and 4 guests