Home | Community | Message Board
You are not signed in.
Sign In New Account
Forum Index Search Posts Trusted Vendors Highlights Galleries FAQ User List Chat Store Random Growery

Kratom Eye
This site includes paid links. Please support our sponsors.

General Interest > Philosophy, Sociology & Psychology
Threaded Mode Threaded  Previous topic Previous  View all topics Index  Next topic Next 

Welcome to the Shroomery Message Board! You are experiencing a small sample of what the site has to offer. Please login or register to post messages and view our exclusive members-only content. You'll gain access to additional forums, file attachments, board customizations, encrypted private messages, and much more!

Myyco.com Shop: Golden Teacher Liquid Culture For Sale

Jump to first unread post Pages: | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |  [ show all ]
Re: "Artificial Intelligence" [Re: Vegas]
    #5034178 -

You can't seriously look at video games to see the most cutting-edge AI around :smirk:

Video games have notoriously shitty AI in them, and although it has gotten better over the years...it's nowhere near to what it could be right now.

Video game designers are more interested in having their cut-scenes and graphics look really nice than they are about AI in the game.


--------------------
Once, men turned their thinking over to machines in the hope that this would set them free.
But that only permitted other men with machines to enslave them.

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: trendal]
    #5034206 -

trendal said:
You can't seriously look at video games to see the most cutting-edge AI around :smirk:

Video games have notoriously shitty AI in them, and although it has gotten better over the years...it's nowhere near to what it could be right now.

Video game designers are more interested in having their cut-scenes and graphics look really nice than they are about AI in the game.


Haha, yea, I kno. Was just responding to what was said basically. I need to read on AI more.


--------------------
"I smoke chronic, pop speed, get drunk, throw up, wake up in the dump, with some of my lunch
covering the front of my sweater, smellin of skunk, deader than fuck and still ever ready for drugs
I won't hesitate to take em, I hate my life, I won't ever make you wait, shit I'll take my pipe
and load it up, with the ripest nugs I sight and puff 'till the night turns to light, I'm the type
to steal prescriptions just to feel so twisted, I can't feel my skin tissue or limbs, witness
my vision's when it's shifting, bless me quick or catch my head spinning like the Exorcist
No, I don't ever wanna come down from this, escape is everything I ever wanted.
This place is where I go to get away from all of this, take me home again."

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Annom]
    #5034212 -

Annom said:
the human brain is a very complex machine/computer


I agree, but one of the main differences between an organic brain and a computer is that computer code operates using functions, while neurons can have multiple outputs.

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Vegas]
    #5034237 -

Intelligence is intelligence. "Artificial" intelligence isn't artificial in that it is functional. It is functioning intelligence. The fact that we have consciously created the mechanisms and their subsequent limitations that bring forth and bound the limited sense of intelligence, it would seem to me that it is still intelligence and that it fufills its purpose as intelligence.

Unless I am not fully understood as to the specific meaning represented by the term "intelligence"..... Nope, I think I got it. The means by which to aquire and effectively apply knowledge? Information, knowledge, itself? :smirk:

Reading through the thread, though, it seems the main issue that this discussion relies on is the considerable distinctions between intelligence and cognition. :lol:

:headbang: :headbang: :headbang: :satansmoking:
Peace. :mushroom2:


--------------------
:redpanda:
If I should die this very moment
I wouldn't fear
For I've never known completeness
Like being here
Wrapped in the warmth of you
Loving every breath of you
:heartpump: :bunnyhug: :yinyang:

:yinyang: :levitate: :earth: :levitate: :yinyang:

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: raytrace]
    #5034524 -

so... the point your trying to make is that they wont succeed at creating artificial intelligence?


:tongue:


k bye.


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


Everything I post is fiction.

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Moonshoe]
    #5034801 -

but we will make more self awareness directed towards the improvement of experiencing.
call it artificial if you like
mine already is


--------------------
:brainfart: _ :finger:

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: raytrace]
    #5034811 -

Okay, so what I'm getting as the gist of this thread:

1) We have no proper definition for "artificial intelligence"
2) ...but whatever it is, it won't ever be made.

Which leads me to the following conclusion:

Raytrace hasn't got a clue what "artificial intelligence" is, yet he is trying to deny the possibility of it ever being built.  :smirk:

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: GulGen]
    #5035191 -

i work a afb in sacramento that was front runner in bs.
ai first spotlight to us from ms in the from of a
10 page insert they bundled with c. we xerox the cover page
and write some bullshit, we need more money from general.

well the big guy was really impressed w/ its version 99 thousand
sofware up grade to the f-111 and the af figured that thats
waht me must be takling abt too. little did he know...

we sold him a bill of goods that ms would be proud of


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

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: rwilber]
    #5035318 -

^ its a whole new breed of crazy ^


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


Everything I post is fiction.

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: trendal]
    #5037871 -

> What about infants?

I read a very cool paper last night about prenatal learning in the visual cortex. The basic idea is that a fetus spends most of its time in REM. They think that during this time, the developing brain is actually learning based on the brain waves generated during REM.


--------------------
Just another spore in the wind.

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Seuss]
    #5037972 -

Quote:
Seuss, that's really interesting. I'm not a programmer, but I did follow a decent Java course for my study, is there any way I could understand the basics of the code of a simple evolution program you wrote?


It has been a while since I have worked with this, but I will try to find my old code. The games I mostly work with are tic-tact-toe, connect four, and Othello. Tic-tac-toe is a very simple game and takes very little time for a solution to evolve. This is a good way to test that a network is learning properly. Connect four builds upon tic-tac-toe increasing the board size and the goal size. Othello builds again, increasing the complexity of the game yet again.

The heart of the code I wrote is based around a self-organizing map (SOM). This is a form of unsupervised learning. For example, it takes a newborn baby around a week to learn how to focus it's eyes. The newborn learns this on it's own, not with somebody teaching it how to focus. A self-organizing map works the same way. It learns to distinguish patterns based upon the patterns that it has previously seen.

The SOM is implemented with a feed-forward neural network. The network shape is defined by an artificial DNA... just a long string of numerical values. Rather than try to encode all the weights and connections, I encode functions that will build the neurons and connections. When the DNA is expressed, a finite set of SOMs are grown. The outputs from one SOM become the inputs to others. Each neuron decides who it gets input from, based upon the DNA coded function for that neuron. Weights are initially set to -1 or 1 (inhibit or excite) and are normalized to unity. Again, and DNA coded function determines the connections and the initial weight (-1 or 1).

A population of organisms is created. Initially, the population watches a computer program play the game. During this time, the SOM is learning the pattern (rules) of the game and play. As the organism ages, it is eventually allowed to compete in the games. As the organisms play each other, their fitness is recorded based upon how well they play. The fitness function is also scaled by the age of the organism. The allows young organisms to survive before they learn how to play and rewards older organisms that have not been killed off as well.

After some set length of time (number of games played), the population is culled by fitness. The least fit of the population are deleted, and the remaining population is allowed to 'mate' until the population size is back to normal. I have used many different mating rules. My current implementation actually mixes up several different strategies... some blocks of DNA are averaged together, other times the section from only one parent is used, other times random crosses occur, and yet other times random mutations are inserted.

I tend to run several populations simultaneously. Organisms from each distinct population are crossed from time to time to help prevent local minimums and help maintain genetic diversity. Children in the population are never culled, allowing a mutation to find its way into the population even if the overall fitness of that mutated organism is poor.

Inputs and outputs into the system are treated as bits in a number. For an Othello board, I have 128 inputs (two neurons for each tile). A value of 00 or 11 means the tile is empty. A value of 10 means the tile is white while a value of 01 means the tile is black. The output for Othello is a 64-bit value with each bit representing a single tile on the board. The bit with the highest value (the most excited neuron) is the play that is made. (I use the word 'bit' loosely here... implementation is a floating point number normalized between -1 and 1...) Sometimes I simply ignore (mask out) all the bits that are not valid moves while other times I end the game as a loss (depending upon the experiment that I am doing).

I also use a decaying timeout after a neuron fires limiting the fire rate of the neuron. Again, this is determined on a neuron by neuron basis by a DNA defined function. Once a neuron fires, it will not be able to fire again for so many 'clock' ticks. This allows the system to create state machines and timers more easily.

I also assign inputs to a timer that represents the percentage of time left in game play. An output neuron acts to signal that the current state of the network should be used for a move. Rather than cycle the network for a fixed number of clock tics, I allow the network to think as long as it likes for each move. The organism must learn to manage it's own game time.

Another adaptation I have made on the above is using the network as a board evaluation function only within an AB-minimax search tree. In these cases, I use many multiple networks. One as the evaluation function, one as a predictor for the opponent's move, one to sort the moves at each node from most likely best move to least likely best move, one to decide time management, etc.

The networks are constantly learning while they play through the self-organizing model. This means that the organism (usually) gets better at the game as it lives longer. However, this learned knowledge is not passed on to the children. Only the design of the network is passed, not what the network learned. This tends to produce organisms that can quickly learn how to play the game rather than organisms that are born with instinctual knowledge of how to play the game. The entire population of organisms that are not currently playing 'watch' the organisms that are playing. During this time, the watching organisms are still modifying their networks, but their fitness functions are not altered.


--------------------
Just another spore in the wind.

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Seuss]
    #5038906 -

Seuss said:
> What about infants?

I read a very cool paper last night about prenatal learning in the visual cortex.    The basic idea is that a fetus spends most of its time in REM.  They think that during this time, the developing brain is actually learning based on the brain waves generated during REM.


Interesting! Sort of like the POST on a pc, I guess :wink:

I had a discussion with Neuro on IRC a month or two ago, were we compared the human brain to a modern computer.

I suggested that all humans will have something equivalent to a ROM BIOS in their brains - certain sets of instructions that are hard-wired into the human brain right from birth. The "OS" then would be continuously programmed during life as new experience/information comes in - it would be the person's personality. You can always change your OS...but you generally can't change your BIOS without getting a whole new system :wink:


--------------------
Once, men turned their thinking over to machines in the hope that this would set them free.
But that only permitted other men with machines to enslave them.

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Seuss]
    #5043027 -

Thanks for your reply Seuss!

I had to read it twice, but I get the basic idea of the program.

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Annom]
    #5043072 -

I once created an attitude media processor
sound in >>
record & analyse
speech recognizer >>
attitude context processor with error recovery reinterpreter
auto selection of conditioned branches >>
expression output by multimedia like max headroom

each expression comes with it's own set of context recognition and error processing branches.

if speech has low recognition in the current attitude block,
previous blocks are used to find a better branch.


essentially it could hold pretty realistic conversations with shifts and previous context reversals.
but it was just basic attitude expresssion and response, needs further development.

wehn it becomes fully conditionalble (learning) has attitude, and can shift mental state (emotions/getting stoned) there will be no difference except from us in body type.


--------------------
:brainfart: _ :finger:

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: redgreenvines]
    #5043112 -

Then you wil be the first person, to successfully create a computer that actually passes the Turing test. That would certainly be an historic achievement indeed. :smirk:


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

Edited by Sinbad (12/10/05 08:28 AM)

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Sinbad]
    #5043178 -

I think it will happen sooner than expected.

New brain scanning results indicate the brain is a complex parallel processor which is malleable and shaped from thoughts interacting with expirences.


--------------------
As the life of a candle,
my wick will burn out.
But, the fire of my mind
shall beam into infinite.

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: psyka]
    #5043230 -

Do you have any sources for these new brain scanning results? As i have been taught, by my university lecturer's, that all of these ideas are just pure speculation that have yet to be thoroughly tested to the point of being scientifically proven.


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

Edited by Sinbad (12/10/05 11:34 AM)

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Annom]
    #5097222 -

From my materialistic point of view, the human brain is a very complex machine/computer
What do you mean from your ?point of view?? Is the synthetic brain going to be conscious only for you and those adhering to your ?point of view??

Is the human brain objectively a computer, or not?

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: psyka]
    #5097223 -

I think it will happen sooner than expected.
Why is that so? Please provide us with something to chew.

And it is expected when exactly?

Extras: Filter Print Post Top
Re: "Artificial Intelligence" [Re: Seuss]
    #5097282 -

If our bodies are basically machines or vessals and we reached conciouness, why cant man-made machines reach this conciounsess?

Extras: Filter Print Post Top
Jump to top Pages: | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |  [ show all ]

Myyco.com Shop: Golden Teacher Liquid Culture For Sale

General Interest > Philosophy, Sociology & Psychology
Threaded Mode Threaded  Previous topic Previous  View all topics Index  Next topic Next 

Similar ThreadsPosterViewsRepliesLast post
* Artificial Intelligence and Free Will
( 1 2 3 all )		 xnevermore 3,955 41 11/25/02 12:39 AM
by Anonymous
* On Artificial Intelligence Catalysis 1,575 8 04/25/02 12:41 PM
by Revelation
* (TT) time travel mapping, (AI) artificial intelligence and random event generator hardware
( 1 2 all )		 Tag_Number 4,031 29 11/20/04 02:07 AM
by Tag_Number
* AI: the simulation objection Sclorch 1,645 10 10/20/02 09:16 AM
by Nomad
* Virtual Intelligence pattern 688 5 12/06/02 06:33 AM
by raytrace
* prove your intelligence: define intelligence!
( 1 2 all )		 atomikfunksoldier 2,843 30 04/26/03 02:20 PM
by atomikfunksoldier
* For those of you who see mushrooms as sacred
( 1 2 3 4 5 6 7 all )		 Colonel Kurtz Ph.D 13,637 133 07/12/05 10:37 PM
by GNIOM1498
* Reality: Our objective, benelovent friend SkorpivoMusterion 1,348 10 11/19/05 03:04 PM
by Ped

Extra information
You cannot start new topics / You cannot reply to topics
HTML is disabled / BBCode is enabled
Moderator: Middleman, DividedQuantum
9,659 topic views. 0 members, 1 guests and 35 web crawlers are browsing this forum.
[ Show Images Only | Sort by Score | Print Topic ]
Search this thread:

Copyright 1997-2026 Mind Media. Some rights reserved.

Generated in 0.03 seconds spending 0.005 seconds on 16 queries.