Thread: Emergent behaviours of the Internet

The internet has expanded to an enormous size, now with over 5 billion connected devices. Ranging from computers, mobile phones, consoles, even refridgerators and cars. More and more of internet traffic is becoming automated, with agent programs performing searches, generating spam email, chat bots, scam bots, maintenence scanning and updating hardware systems and countless other applications.

The internet has also shown a number of emergent properties in self similarity, reinforcing common traffic routes, and some self organization in traffic congestion.
We also have advances in Artificial intelligence (albeit primitive) and distributed computing, viruses that can "learn" and modify and evolve its own code, and exist within the internet independantly of human input.

With all this in mind, do you believe it is possible for the internet to eventually display behaviour that may be interpreted as intelligent to a certain degree? perhaps not as a single entity, but to the point where it is not just a one way tool for us humans. In a way it can interact with the external world, in much the same we we interact with it, and not just be some kind of meta-world within electrical signals.

I am not suggesting something like skynet with a distributed computing "destroy the world" virus. But merely the countless automated and autonomous "wild" programs interacting in ways unanticipated. Im just curious as to where the progression of the internet network is leading. Allready I receive "spoetry" in my inbox which seems like a rambling child learning to write.

Maybe in a few hundred years it may be the liberator of our mortal biology.

What are your thoughts?

Brilliant thought :-D.

That would be cool to see, what if all of the 'intelligent viruses' and their self presevation developed on its own, and grew from more and more use of cyberspace to then transcend itself into every computer connected to the Internet.

What if that intelligence and self preservation become sentient, by creating a self replicating virus that communicates with other computers and updates the virus much faster than humans can eliminate it?

What I wonder is if eventually it would become a menace virus, albiet aggressive and manipulative, or if for its survival it grew into the background of a computers' background processes?

Harvestein, how did this idea pop into your head? (love the drawing of yourself too btw)

I had been pondering it for a while. I started to wonder just what the internet would be like in say 50 years or a hundred years from the way things are now.
I think there are too many parallels in the mechanisms of natural life/evolution and the structure and evolution of the internet (and its contents) for it to be ignored. Its clear that actually many AI programs are based on what we think we know about our own processes. "How the mind works" by Steven Pinker describes it quite clearly.

The total computation power of all connected devices must be astronomical. A worm/virus that could harness that in some way would be very powerful indeed, and surely that would be a natural goal for any self-evolving program. To spread and maintain coherence/communication with its kin. I think bot nets allready do this to a certain extent. That could lead to all kinds of interesting possibilities. Also give it evolutionary advantage over the Norton AV agents sent out to get them. ha.

on a more out-there note:
I just finished a book this week called "Matter" by Ian M. Banks. Generally I did not think it was very good. But it described a fictional society called "the culture", where sentience had become something that was easily moved around or transferred. For example the majority of people in this society could modify their biology as they wished, and more advanced people in that society existed entirely within something called a "dataverse". Which I think is the writers vision of the internet of the future. Sentient intelligence can be uploaded to the internet and exist within it as eaily as a jpeg or a .mp3. Natural intelligence and AI co-exist.
"Normal" members of this society could also hook in or out of this dataverse as they wished without the use of any peripherals. Kind of like telepathic wi-fi.

I think he could have written a book entirely about that, rather than have it as a curiosity amongst a fairly poor storyline.

Originally Posted by harvestein

do you believe it is possible for the internet to eventually display behaviour that may be interpreted as intelligent to a certain degree?

While the notion of evolutionary progress, from lower to higher, has gone out of vogue, I think we may still call the parallel succession of emergent entities "progress". Why would it stop?

But I doubt anything emerging as, say, the Internet, would display behavior appreciated by us. The nature of emergence is such that its ...minions... cannot grasp it. For examples your cells have no awareness of your body, and parts suppliers can't always say if they're making an alarm clock or TV or toaster. Stockbrokers needn't fathom what they're doing - and perhaps it's better when they don't!

I think it begins, to our perspective, with finding life a little easier by doing something a certain way. We begin to depend upon it. Eventually we find ourselves serving that which was supposed to serve us.

If you're the type who resents bloated bureaucracy or having to drive a car to buy an apple, you won't like what the Internet has in store for us.

Several media references come to mind - movies, shows, books, etc. But the concept of transhumanism and the technological singularity also strike me by reading this topic.

http://humanityplus.org/learn/transh...t-declaration/

It's a pretty cool (if not scary) thought to think of a computer virus gaining sentience.

Here's a thought that may be of some concern. I think most of us are aware of progress being made to communicate with computers with the mind directly. To me that implies that the computer will have access to peoples brains.

On the surface that doesn't sound to bad and might be very good. Every time you think of something you get instant feedback based on the most current information the human race has stored on the Internet. In health care alone that would make an unbelievable difference, and there would never be anybody in ignorance of the law. An from the legal systems point of view, you can't lie to a computer that can read your mind, so crime would become a thing of the past.

But let's hope no one creates a virus to take advantage of that wonderful system of feedback.

With all this in mind, do you believe it is possible for the internet to eventually display behaviour that may be interpreted as intelligent to a certain degree?

It brings chill to my spine when I think about self programming computers .

Originally Posted by Hippocampus

It brings chill to my spine when I think about self programming computers .

Is that a good chill or a bad chill? Details please?

Originally Posted by Lance Wenban

Originally Posted by Hippocampus

It brings chill to my spine when I think about self programming computers .

Is that a good chill or a bad chill? Details please?

Bad, unless there was a ‘world police’ guarding that those ‘new things’ will represent our ethics and no evil computer nerds can interfere with the future software.

I am already tired how the laws of evolution and natural selection determine the way I act and how they influence my ‘moral spine’.

Originally Posted by Hippocampus

Bad, unless there was a ‘world police’ guarding that those ‘new things’ will represent our ethics and no evil computer nerds can interfere with the future software.

I am already tired how the laws of evolution and natural selection determine the way I act and how they influence my ‘moral spine’.

Did you ever think that humans might be destined to create the next step in the evolution of life? Life evolved beyond biology. After all self aware intelligent machines could never evolve without the help of biological intelligence.

The human race is facing many extinction challenges and the one thing that will help us avoid extinction is by not having all of our eggs in one basket. In other words we need more worlds besides the Earth to live on. If we had all the time we need to do that, I'd say no problem. The question is do we have that kind of time. Now I know with a little thought I could list a couple of dozen things that have a possibility of happening in the next 100 years that would practically guarantee we don't make it.

Having the help of self aware super intelligent machines would speed up our ability to start finding more baskets to put our eggs in. So I believe the benefits out way the risks of creating machine intelligence.

Also I might add that there is nothing obvious I can see that theoretically prevents our evolutionary path merging with the electronic one we have created.

It is allready a medium that we use to interact with each other, and interact with a vast vault of accumulated knowledge. I don't think it is a massive leap to assume that it may be implemented in such a way that the line between what is distinctly human and virtual can become blurred, much like the idea proposed in the book I mentioned. A 3G internet device interfaced with the brain for example. Sounds very sci-fi, but you never know. There have been experiments getting cat and mouse brains to interact with neural>electronic interfaces. Albeit very primitive.

I don't see it as a threat, as in Terminator style threat. I am concerned about the consequences for privacy and individuality though. Where will the person end and the "cloud" begin?

On a side note, there is a bit of news that may interest to you guys:

http://www.bbc.co.uk/news/technology-12400647

I always thought the biggest obstacle for using the internet as a learning tool for AI was getting it to understand the laberynth that is the net and human language. But now moves are being made to standardise and create a "robot" internet, so that any learning machine has a resource to tap into and contribute. Potentially allowing all connected devices to act as the senses of a learning system getting to grips with the world. Would be interesting to see how this develops, especially if it reaches a significant scale of usage. The protocols used would have to be very flexible though to allow diverse learning, and not just be "GPS location xxx contains an obstacle".

this reminds me of the Gaia theory of the earth as a giant biosphere/organism - perhaps the internet is its neurological system

well... internet make people go riot & do revolution... doesn't this shows that internet control humans!? it even control who is elected president and what law to pass or not to pass. Amazing?

That is an emergent behaviour alright...

From one perspective, all of us exchanging information here and right now are not the cells and flesh we perceive ourselves to be, we are the interaction between these cells (and between the environment and these cells).
The cells dont know whats going on, they dont even know they are cells, and dont know we are made of them (or of their interaction from one perspective), they dont know we exist so cant communicate with us (the meta consciousness) on an individual level and we cant communicate(generally speaking) with them. Our consciousness has emerged while non of the cells ever knew about it, and for most of our existence as sentient entities we didnt know they even existed(and had no clue what we were, nor what we really were). Because the cells arent self aware we have no means to communicate with cells about our reality.

If we are eventually part of a meta consciousness, this meta consciousness might initially not know it exists and go throught a learning process(like an infant that doesnt know what are the hands and doesnt yet know hes moving them or hurting himself unintentionally), its perception of reality might also be very alien to us, the same way individual retina cells triggering signals have no clue what a red apple is or that they are relating that information (as interpreted by the outer-dimensional entity we are).

So yes I think its possible for the internet(that is all systems and the neural nodes which are the users brains) could become intelligent or self aware, but it might not act intelligently as it learns and becomes aware, and it may not be able to relate/communicate with us as individuals nor perceive the world in a manner we can easily relate to. As it learns, we probably would not perceive or realize that events from our reality's perspective are the actions be it controlled or not, of this meta entity we would be a part of. On the plus side we are self aware so my guess is its easier to communicate with us, than it would be for us to communicate with a hair cell.

(Or it might be a conventional artificial intelligence that writes on some computer screen "hello?")

Absolutely agree. ie: Our neuron might be part of the brain, but the brain itself is vastly different than the individual neurons and it is impossible to imagine how the brain (the larger entity) could communicate with the neuron (the subset) at the same level. For sure an individual person might be like a cell: which can't possible aware that it is part of a brain entity.