The Creative Spark

[Bimboliquido]

Curious how everyone ignored my post while they answered to Mu who just said what i said in two lines...
Nice forum,this one!

[Jeni Nielsen]

Curious how everyone ignored my post while they answered to Mu who just said what i said in two lines...
Nice forum,this one!

Er well if you really want feedback make your post easier to read. I went back and tried to read your post and it wasn't easy to do.

Granted it doesn't have to be perfect or anything, but paragraphs and punctuation are always good ways to get people to want to read what you have to say

[Bimboliquido]

Er well if you really want feedback make your post easier to read. I went back and tried to read your post and it wasn't easy to do.

Granted it doesn't have to be perfect or anything, but paragraphs and punctuation are always good ways to get people to want to read what you have to say

Can't say you are wrong,but i'll admit that given the reading difficulty of the first posts i thought that this wasn't an obstacle.Well,my mistake,and after all,who cares?
See ya when i'll learn a little more written english.

[Lethagrin]

AI will never surpass us because they need us to reproduce them. And they are not biological creatures so they can't evolve and adapt when nessesary. they also need us to mantane them.

Well I don't know. Think about how cars get built nowadays. It's mostly machines building other machines. Who's to say that it won't end up being a completely automated process in the future. So in essence machines will be able to "reproduce" themselves.

Indeed. And besides, technology evolves. The technological evolution is both faster and more efficent than biological evolution. When a human-level AI is realized, it will be able to take part in designing a super-human AI, which alone can design AIs of greater and greater capability. It is evolution, but unlike our evolution, it's guided by intelligence and proceeds in matter of months or even days, rather than millions of years.
No human interference is required. There is no reason, why a machine of suitable complexity could not possess all creativity of human mind and more.

Will humans be replaced? I don't think so. After all, like said, the AIs will be created by us and raised by us. Essentially, they will become us, though far more intelligent and capable.

Bringing this post back to topic...
Let’s not forget that the standard meaning of "evolution", at least in an organism perspective, means survival of the fittest due to genetic mutations to haploid sex cells (sperm and egg) created prior to fertilization. The traits inherited by future generations are used as a strength (or weakness) in their environment. These aspects of evolving can only happen in a society level not an individual.

The point I'm making is if AIs are (or will be) considered a type of organism or something close to "living" then it would first need to have sex in order to create variation and "evolve" on its own and it would also need to have random changes in the "code" it passes on to future generations. If it doesn’t do this or it doesn’t even "create" other entities then it’s not evolving its adapting. Seeing this, then wouldn’t it still just be something which was “told” what to do rather than acting independently?

Would AI's be made to be able to "create" new individualities and if so how would the process undergo? Also, would there be a need for new individualities seeing merely one could sustain itself for an eternity?

Would there even be a point to creating new identities, besides the reason of doing it to do it, if a single one could adapt just as much as a the one it was duplicated from?

I'm not saying that AIs have to be in the reflection of humans, however, man kind's creations tipically represent themselves more than anything.

The golden apple doesn't fall far from the tree.

[Lightice]

Bringing this post back to topic...

I'm not sure if the forum rules consider this necromancy or not, but new discussion is always welcome, as far as I am concerned.

Let’s not forget that the standard meaning of "evolution", at least in an organism perspective, means survival of the fittest due to genetic mutations to haploid sex cells (sperm and egg) created prior to fertilization. The traits inherited by future generations are used as a strength (or weakness) in their environment. These aspects of evolving can only happen in a society level not an individual.

Actually reproduction with sex is relatively new idea, from the viewpoint of life in general. Sexual reproduction is the ruling paradigm among the most advanced species on the planet, at the moment, but that doesn't mean it's the best possible method available. I can think at least one better alternative viable to AIs, or more accurately, inorganic humans - as long as there is real division between human and an artifical intelligence, we haven't gotten far enough.

An AI can consiously design new artifical entities, like we can design AIs. Each generation is more advanced than the last and thanks to the consious design-process instead of blind evolution, the speed of advancement will in a completely different league. Random process isn't required - indeed, it's not even desireable, as long as the formation of the personality is left free - for a given value of free, since it's not like we are entirely free to grow as we like, either.

I'm not saying that AIs have to be in the reflection of humans, however, man kind's creations tipically represent themselves more than anything.

I think AIs should be a new model of humans, in order to prevent them being just advanced tools.

[Lethagrin]

Excuse me for making it look like I was refering my quote when I said "Bringing this post back to topic..."
That was not what I intended. I was refering to the previous post to mine. The one talking about post length and such. I infact thought that the posts I quoted were very interesting and I wanted to shine light on them. Also to open that part of the topic again.

If an AI doesn't "evolve" with random change then what would it resolve to? Trial and error?

[Jeni Nielsen]

Bringing this post back to topic...

I'm not sure if the forum rules consider this necromancy or not, but new discussion is always welcome, as far as I am concerned.

OTing the post again...

There's no rule specifically saying what you can bring back as a post or not. I'm fine with the resurrection of any old post, provided people have something new to say.

/OT

[Lightice]

If an AI doesn't "evolve" with random change then what would it resolve to? Trial and error?

Presumably to observation and simulation. If it can study its own source code, it should be able to improve on it.

[Epiphany]

AI will never surpass us because they need us to reproduce them. And they are not biological creatures so they can't evolve and adapt when nessesary. they also need us to mantane them.

I think AI's will pass us by. As pointed out the manufacturing industry is already making people useless. Computers of a year ago are already obsolete. Plus somewhere I read that MIT already have a computer that can physically interface with a human brain. They are working on computer controlled prostetics. The famous 3-laws have already been proven to be flawed to the point of being the main reason for our downfall. Our military is working on machines that can select and attack targets without any human input.

Plus killer AI's make good movies

[sonic]

And if AI develop emotion I hope they hav compassion becuase if they don't than that would suck.

You mean the kind of compassion that humans show their "inferior" species, such as the other life on this planet? Let's hope that there aren't as many people willing to put cows unnecessarily through the chopper and okay with thinking of other creatures as nothing more than resources to be farmed... Let's hope the AIs don't behave as the majority of humanity does in that respect. Humans put their closest cousins the apes in lab tests, though- does man act in a way that seems undeserving of similar treatment from a higher lifeform than it? (Of course many individuals do better, but fate doesn't exactly kindly stop and look and say. "Oh, but look at the individuals")

On a side note, looking at the posts again I feel kind of sorry for the Italian lady, even though she got a bit snippy. I think she didn't realise that people were quoting Mu in their post because he gave a succinct couple of lines that were easier to quote and reply to, and didn't feel the need to reply directly to her one before it because they were then talking about things along those lines anyway. Also I think she mistook the response for an attack. Oh well, a pity...

At any rate I do not really think AIs will behave in a sinister or evil fashion (by our standards). They are evolution, and my positive nature means I see the best for that. People getting afraid speaks more of man's fears, vanity and self-importance than anything. Of course we should be prepared and try to protect ourselves just in case it's the worse case scenario, but really I don't see the potential developments going badly unless it's man's fault for doing something screwy at the outset- like in Bubblegum Crisis for instance where the cold-hearted Genom corporation created sentient robot girls and used them as slaves, so the girls turn on their masters when they can't stand the way they're treated any longer, and they end up killing some of them and some innocent humans too (though only in their desire to be free).

It's a bit like humans and dogs really. Humans acting out of their worser nature breed dogs for fighting. They push their luck with them and then is it any wonder one day when the unstable creature they've created turns that killer nature on them? Humans who try to segregate, try to make themselves superior, try to establish their mastery over something with a firm and cruel hand, try to make advancements acting out of their own selfish worser needs; those humans will face the products of their own choices one day, and the way it has destroyed and corrupted them. Unfortunately, the whole of the human race bears that weight. After all, the killer dog you beat until it snapped and became a monster isn't just gonna go after you- it doesn't even notice or discriminate whether it's the bad men or the small child in the street anymore...

As an aside, my favourite character of my own creation is a nurturing AI-born person who has faith in humanity and in their own humanity and wishes to help guide the world positively. They would be very keen on experiencing life first hand, especially after reading everything there is to know about it on the net, and like Star Trek's Data they are into being more human; strengthening and affirming their own personhood while still retaining the grace and wisdom of someone who isn't us and never can truly be one of us; as somthing evolved beyond us. I'd prefer a character who wasn't all about the dark side of the AI debate. They simply have love for the world and what/who can be in it, and they wish to protect it and be a part of it, and to emancipate others (i.e. humans) that they see into striving for the best for the world too. It's kind of the most positive view of the future I can think of; a future you can dream for. Sort of like cyberesque sci-fi meets shoujo I guess, in terms of feel and outlook. It's the kind of positive story that I as a person would want to write and give to the world, it's very much my personality sort of story. Fight to show love to the world and to evolve, and all that. Don't act like an isolated god- take others with you, and help them help themselves and the world around them. This is the best case scenario if AIs become what could be considered people, and it would be wonderful. And the beauty of it is that we humans helped make it happen, and still play some active part in shaping the world for the better. I'm an optimist, but I want to give people something to dream about and feel hopeful about... Like the first Ghost in the Shell film, Final Fantasy VIII, and countless other things did for me as a teenager. We can get better- and we might make something wonderful, wonderful for the world, out of love and inspiration one day... (I always loved the end line of Macross Plus- "Dedicated to you, our future pioneers...")

Heh heh, if this sounds weird ignore

[AI]

It's a bit like humans and dogs really. Humans acting out of their worser nature breed dogs for fighting. They push their luck with them and then is it any wonder one day when the unstable creature they've created turns that killer nature on them?

Well said and a good analogy. This could apply to AI's in the sense that an AI developed and contained in a lab would not have a reason to develop a killer nature however an AI embedded in a battle tank that is programmed to have 'killer nature' or to disregard the value of (human) life, would be more of a thereat ( to it's creator) if it's thought process was not under adequate constraints.

[holdenmcclure]

Warning: Incredible wall of text approaching

There are two things that need to be acknowledged before you read my post:
1) I support the theory that the universe was never "Created", rather it has always existed in one form or another (Big Bang).
2) I do not acknowledge the existence of the human soul and attribute all human behaviors to genetic predisposition and environmental influence.

Now then lets begin:
There seems to be a misconception floating around that robots are all metal and gears. It is well within our realms to create synthetic duplicates of the human form. It's only a matter of time and motivation. Human behavior is attributed solely to the electromagnetic signals sent by our brains in order to stimulate our bodies to perform various actions and events. Although it sounds quite short and simple, in the scope of human behavior its a vastly open spectrum for varying humans. When mankind develops the tools, knowledge, and full understanding of the human brain, mankind will be able to duplicate and manipulate everything that humans have used to justify their individuality.

That's not to say that a human brain is required for conscious thought. There is no specific line between unconscious thought and conscious thought. Being aware of one's self falls under a similar category. In programming terms, what humans refer to as conscious thought is a decreasing line while programs are becoming more and more capable to write and execute unique actions. Ultimately a program and a human is both dependent and limited by its scripting or genetic predisposition. The more specific our DNA is, the less variance in our behavior, however that also means that our behavior is more streamlined for the environment. Simple things that Humans developed, such has pose able thumbs have allowed humans to create and manipulate the environment in ways that other animals have yet to achieve, allowing more us to develop tools that render our natural abilities useless, thus eliminating them from out bodies via natural selection.

Understanding that with tools we are capable of recreating and manipulating anything that exists today, and acknowledging the many possibilities and variations of the human form. It's only a matter of time before being human is entirely insignificant. The truth is scary, and it undermines the holy image of mankind that we have eluded ourselves in. We will someday have to acknowledge that mankind's is just another arrangement of molecules, and all the differences we use to emphasize our individuality is merely a stepping stone for all species.

[Aoi]

The problem nobody's talked about is programming creativity. All AI's we have now are narrow, and have come about because of specialization. Yes, they could look for math theorems, "analyze" things, and orient in spatial reality, but these things are horribly limited in the way the machine has any sort of originality.

Consider this: given any possible problem, a human will always have thoughts and ideas about it. But can a program, held only to its code, be capable of looking into any possible problem? Humans work by learning and understanding, not just trial-and-error adaptation and in-built presets.

In SAC, Motoko talks about how our ghosts whisper to us, and she was referring to our inexplicable intuitive reasoning and insight. When we don't know how our brains go about learning and understanding, how can we program for such things?

I don't think just copying the brain's neural networks is going to give us an answer because there's just too much going on in it to analyze. In order to get an AI even up to our own level, we need to know exactly how we function at our own level, and this has yet to be fully known and modeled.

[GhostLine]

Good point Aoi.

Intuition, or "thin-slicing," happen at a level that is nearly unconscious but somehow surfaces to our thoughts. And our intuition often defies logic.

[AI]

The problem nobody's talked about is programming creativity

What if creativity is just a form of logic that is so abstract that it cannot be emulated, that is to say that humans don't understand it so cant willfully create a device to emulate it. However, if an AI were to be made to evolve it's logic processes could possibly acquire this trait.