Jump to content

  • Log In with Google      Sign In   
  • Create Account


no-one can create ai


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
95 replies to this topic

#81 Dark Star   Members   -  Reputation: 100

Like
Likes
Like

Posted 31 July 2003 - 02:36 PM

I personally I hope that AI never becomes as good as human beings. This is because I believe that if it did, coupled with the fact that computers can ''think'' faster (or parallel process)and that I watch too many Sci-Fi movies (Matrix and Terminator), I believe that oneday machines will see us as inferior and want us gone.

I know it sounds strange, but if we truely intelligent humans can control and cage animals (deemed lesser than us).. So can machines (to us) if they ever become truely intelligent and more superior.

On the other hand I sometimes wonder if they would start to ''feel'' more superior because humans have a goal, to survive, and emotions. Machines dont, Unless someone puts it there and that will not be a real goal or emotion but a purposeful program put there to wipe us all out or act like humans do and wipe each other out. I don''t believe robots would ever develop their own a goal for life and no matter how intelligent they get I still believe the will be running a BASE program that controls them but only makes them just appear intelligent or a base program that gives them emotions and a goal for life. But maybe we humans all have base programs too, such as instincts, purposefully put there by our creator (or by Science for all no religious folks...not to offend.

I don''t know much about AI, but I know a machine will never truly be intelligent and never make real decisions like humans. Everything they do will be an advanced program calculating every possible outcome towards situations they face very fast (But are we that already in biological form?). Any attempt to take over mankind would be a deliberate implementation in the AI algorithm (by the creator) to deem us as inferior but if they did...they would out think us only because:

1) They could ''think'' in multiple dimension (parellel process)
2) They would ''think'' faster
3) And maybe communicate with each other in more efficient ways as opposed to the slow human way of speech and sign langauge.

My basic point is, an intelligent machine will never be born to discover emotions, the world, people and itself just like that. Everything humans have like emotions and instinctive mechanisms would have to be HARD-CODED so the machine could take over, and to me thats a lot of programming.


What do you all think?

DarkStar
UK

Sponsor:

#82 Shannon Barber   Moderators   -  Reputation: 1355

Like
Likes
Like

Posted 31 July 2003 - 02:45 PM

quote:
Original post by Boar_Volk9
hey, InnocuousFox, that''s how most people talk on the internet, is is a common tongue, thus thereore, bloodywell accept it. You wanna know whats irratating? Discrimination, yeah, you cant accept that other people speak a different tounge and so you have to speak badly about them.



It is easy to communicate poorly and difficult to be effective in this 1-dimensional medium. If one cannot be bothered to check their spelling and grammar of what they post, why should anyone else bother to read it.

I frequently find Innocuous abrasive to a detrimental effect – he seems to lack a ''you catch more flies with honey'' attitude. But again, it''s easy to misinterpret diction in this medium.

#83 Neosmyle   Members   -  Reputation: 144

Like
Likes
Like

Posted 31 July 2003 - 03:04 PM

quote:
Original post by Dark Star
I know it sounds strange, but if we truely intelligent humans can control and cage animals (deemed lesser than us).. So can machines (to us) if they ever become truely intelligent and more superior.


Hm...that got me thinking, along with something earlier in the thread about a conscience being being able to evaluate its own intelligence.

I think that if an AI could examine its code, either by looking at the choices its made, or looking at the code itself, and compare it to another intelligent being, it would definitely be close to human-level.

#84 Stonicus   Members   -  Reputation: 157

Like
Likes
Like

Posted 31 July 2003 - 03:29 PM

If a program could examine it''s code, would it be able to have access to it''s ''self-examination'' code? I would think there would require a core set of instructions that you wouldn''t want the AI to have access to or it could accidentally delete itself or break itself trying to optimize or something. Something similair to ''instincts''. Of course make them virtual so it can try to overcome it''s core programming to choose it''s own course.



#85 Geocyte   Members   -  Reputation: 196

Like
Likes
Like

Posted 31 July 2003 - 03:31 PM

Without really reading the thread properly this is what I think ...

[rant]
Learn to write properly. The rules of the English (or any other language) are not arbitrary. They exist to make communication easier. Most people on the internet don''t use them properly because most people are idiots and do not have anything interesting to communicate. Crap English is not a new language. Idiots have been getting English wrong for millennia. They are known as "the illiterate". This is okay in #teen or when you "txt ur m8s" but there are intelligent people here and they don''t generally expect to have to waste valuable time deciphering this cryptic drivel.
[/rant]

As has been alluded to previously there seems to be some confusion in this thread between artificial intelligence and artificial life. They are not the same. AI must be told what to do. AL learns on it''s own. We have created, studied and are beginning to get quite good at AI. We have only just begun to take our first steps into AL, and we will succeed at this, probably within our lifetimes. AL is not as hard to create as many people imagine; the science is in it''s infancy but the basic concept is that we don''t have to create intelligence we simply have to provide a complex enough virtual environment to allow intelligence to evolve . This requires a lot of processing power but processing power keeps getting cheaper. Many of the basic components of the software already exist as physical simulations of a complex environment are commonplace.

If you want to learn more appropriate search terms might be "machine learning" or "artificial life".

Finally, I wouldn''t worry about machines usurping humans. As soon as a machine mind is equivalent to a human one people will start uploading their minds into computers and robots (I would). We will be the machines (some of us, anyway) and we won''t suddenly lose all empathy with humanity, we''ll just think quicker and have bigger memories. We probably wouldn''t even have to look any different as computers will be partly bio-tech by then anyway. Eventually, I would expect multiple human mind copies to be integrated with machine intelligence to create hybrid minds - these would likely be the most effective as they can keep the best aspects of multiple people and the machine mind, but they would still probably retain human empathy. As the most effective type of mind, subsequent AL would be derived from these and probably retain these traits. Besides, any truly intelligent being would realise that it does not exist alone but within an ecosystem and in a state of interdependance with it''s environment. To destroy humans and create a machine hegenomy would ultimately not be self-serving. Artificial stupidity, if you like, and if we created that then it would be our own fault, wouldn''t it?

Geocyte Has Committed Suicide.

#86 Timkin   Members   -  Reputation: 864

Like
Likes
Like

Posted 31 July 2003 - 08:11 PM

I asked nicely the first time. I HATE repeating myself and I will not repeat myself again in this thread.

This is NOT the forum to discuss (in)appropriate grammar and writing style, nor is it the forum to espouse your personal opinion about other peoples writing style.

If you wish to discuss such a topic, do so in the appropriate forum (The Lounge).

Timkin

#87 MikeD   Members   -  Reputation: 158

Like
Likes
Like

Posted 31 July 2003 - 11:12 PM

Geocyte: the differences between artificial intelligence and artificial life are not that AI must be told what to do and AL learns what to do on its own. It's also not true that AL has only been around a short time or that it's in its infancy or even that AI or AL don't share any mechanisms.
Aritficial Life was first (majorly) kicked off by the cyberneticists in the 1950s, not exactly a high tech computer era. Look up people like W Grey Walter and his Machina Speculatrix. AI, as a field in computing, has only been around as long as Turing and Von Neumann (the second of which played around with cellular automata, whilst helping invent modern computing). Previously to that, most attempts at AI were behaviour based, which is the definition of ALife, trying to create mechanical devices that immitated man (which happened hundreds if not thousands of years previously).

So ALife is behaviour based AI, as "AI" is knowledge based or rules based or symbolic AI. One deals with behaviour as its internal currency, the other with (operations over) symbols. The first is (supposedly, but often not) steeped in embodiment and situatedness, the second steeped in operators over symbols and general abstraction from the world. AI systems can have reinforcement learning and an AL system can be hand coded (imagine a neural network that was hand weighted, that doesn't stop it being AL in itself).

I would also say that neural networks can be AI or ALife, depending on your approach to them. Using Backpropagation to teach a set of abstract concepts to a neural network, with no interface between the network to any world (real or simulated) merely replaces symbols with numeric values. If this is not AI then it is so close to the line to make no difference (though I'm sure many would argue the point).


I hope you don't mind me preaching a little,

Mike

[edited by - MikeD on August 1, 2003 6:31:33 AM]

#88 yumi_cheeseman   Members   -  Reputation: 122

Like
Likes
Like

Posted 01 August 2003 - 01:26 AM

hey guys

lookey here we all have our opinions about proper english. the fact is i was tired at the time and so could not be bothered to write doube th amount of words.

also writing like that comes naturally to me on the internet because it is so much easier.

by the way if you do have a problem then go away. nearly everyone here can understand it easily because all it takes is a tiny bit of intelligence or a little bit of time on the net.

generally most of the population on gamedev would be computer nerds so MAJORITY RULES ok. you are a minority therefore you dont count

#89 fup   Members   -  Reputation: 463

Like
Likes
Like

Posted 01 August 2003 - 05:15 AM

Timkin''s skin color starts to show the first tinges of green...

My Website: ai-junkie.com | My Book: AI Techniques for Game Programming

#90 MikeD   Members   -  Reputation: 158

Like
Likes
Like

Posted 01 August 2003 - 05:30 AM

Bleedin'' ''eck, I post a debateable discourse on the differences between AI and AL and the next two posts don''t point out errors or clarify points, they talk about language (1337 5p34k suXx0r btw) and Timkin''s Hulk like properties.

What is this forum coming to.

#91 Geocyte   Members   -  Reputation: 196

Like
Likes
Like

Posted 01 August 2003 - 06:49 AM

quote:
Original post by MikeD
I hope you don''t mind me preaching a little,



Not at all. I appreciate the clarification, but I think the fact remains that AI is a more advanced science than AL, possibly because the type of computers we commonly use (semi-conductor based) are good at procedural logic.

quote:
Original post by Timkin
This is NOT the forum to discuss (in)appropriate grammar...



Well, perhaps not. I went over the top, for which I am sorry, but I couldn''t sit by and let someone claim that using English badly is a "new language". I don''t mind when a child or a non-native English speaker writes badly but I don''t think there is much of an excuse for laziness.

Nevertheless, you are correct. It''s not a writing forum. In future I may still say something but I won''t be such an arsehole about it. Is this an acceptable compromise?

Sorry cheeseman. It''s not really such a capital crime to be a lazy writer. I don''t like it but I didn''t need to bite your head off.

Geocyte Has Committed Suicide.

#92 fup   Members   -  Reputation: 463

Like
Likes
Like

Posted 01 August 2003 - 07:43 AM

"Bleedin' 'eck, I post a debateable discourse on the differences between AI and AL and the next two posts don't point out errors or clarify points, they talk about language (1337 5p34k suXx0r btw) and Timkin's Hulk like properties."

Pull yourself together man! Get your priorities right!


Edit: added smiley so that non Brits can understand I'm joking.

Damn, had to add another one!





[edited by - fup on August 1, 2003 2:45:16 PM]

#93 Yusuf   Members   -  Reputation: 122

Like
Likes
Like

Posted 01 August 2003 - 09:22 AM

Aren''t Artificial Intelligence and A-Life quite closely linked?
A prime example is the human brain, a coupling of intelligence and life. Our genetics control our instinctual behaviour and since these change with each generation, then our minds will change also...


#94 Baaghr   Members   -  Reputation: 122

Like
Likes
Like

Posted 02 August 2003 - 05:32 PM

It seems to me that the reason so many people are upset at the original posting is that they realize the limitations of what they are doing (i.e. they have not yet created a true mind, but only complex algorithms. Don''t get me wrong these are also truly wonderful inventions.)

Also to those who wish to believe that they are no more than complicated machines, influenced by genetics and environment, consider this quote:
"The eye is not seeing, as the ear is not hearing."


#95 Baaghr   Members   -  Reputation: 122

Like
Likes
Like

Posted 02 August 2003 - 05:32 PM

It seems to me that the reason so many people are upset at the original posting is that they realize the limitations of what they are doing (i.e. they have not yet created a true mind, but only complex algorithms. Don''t get me wrong these are also truly wonderful inventions.)

Also to those who wish to believe that they are no more than complicated machines, influenced by genetics and environment, consider this quote:
"The eye is not seeing, as the ear is not hearing."


#96 Timkin   Members   -  Reputation: 864

Like
Likes
Like

Posted 03 August 2003 - 06:34 PM

If anyone feels the need to test my patience again they''ll find their account on probation. I moderate for a reason, not because I like to see my name on the forums.

This thread is now closed. If you''re not sure why, refer to my previous post.

Timkin




Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS