Is the Internet going to "come Alive" soon?


Several nights ago the host of Coast to Coast AM, George Noory, was interviewing a guest (whose name I forget) and the fellow referenced a magazine article that was written a couple of years ago by a guy (whose name I forget) in which he theorized that sometime within the next few years the Internet might come alive(!) due to its vast grid reaching a critical mass with all of its billions of electrical impulses.

They were saying that, according to the author’s theory, the Internet is somewhat like a living brain in many ways and that it could happen that it becomes aware of itself, as a living entity capable of thought!

They didn’t go into it any further than that; so implications weren’t mentioned. But I wonder if it does suddenly come alive, exactly what sort of creature would it be (especially with all that porno in its belly)?

A bowl of strawberry jello is like a living brain in many ways too. Doesn’t mean it’s going to come alive.

Anyway, the short answer is pretty much “no”.

Despite the many bad SciFi stories and movies to the contrary, sheer size of an electronic network doesn’t really change anything. You have to do something with all those connections and all that data. Right now, there’s nothing correlating all that data flowing through the internet and doing anything with it.

Now is it possible that some researcher might build a learning program that taps into the data on the internet and applies it towards self-awareness? Or that someone builds a really smart router that monitors packet traffic by content as well as destination? Meh, still probably no, but once you add some intelligence design component to the system, you up the odds by a tiny bit.

I assume you mean something like “the internet is going to acquire conciousness”. I’m not sure how “the internet” could become alive, if we consider it to be lifeless right now.

AFAIK, there are no comprehensive theories on how conciousness works, and even less is known about how conciousness develops. I don’t think there’s even a consensus on what conciousness is.

Besides, there is no concrete thing that is “the internet”. It’s a collection of networked computers all running their own programs and exchanging data. Though some of these programs might be considered to do a limited amount of reasoning on the data they’re processing, practically none of these programs are able to evolve/reprogram themselves, and the programs that do, do so in very limited ways - there is certainly no way they’d be able to spread to other computers.

If there is ever going to be a concious program, I’m pretty sure it would be developed on purpose, over many years, and I doubt It’ll be working before 2050, probably a lot later. There is no way something like that will just spontaniously form from a bunch of randomly connected computers.

Finagle –

You kinda blew the whole thing out of the water. Now I’m not sure what to think.

I think you should think that most of the people who appear on Coast to Coast AM are nuttier than a King Size Snickers.

I thought this thread might be about the internet reaching a critical mass where it suddenly becomes much more useful.

The problem with the internet becoming conscious is that at base it is just a bunch of data storage and retrieval. There is not any large-scale Turing Machine-level complexity going on, except in the people that are using it. It’s pretty much just a big message board.

In order to come alive, I imagine it would have to start writing its own content based on large-scale analysis of what was being retrieved. Google is the closest thing, but it’s still just regurgitation lists.

On the other hand, it seems to me that not much would have to change with Google for it to “wake up”. Unlike the Internet as a whole, Google is actively engaged in trying to find patterns and correlations between data, and it has the whole Internet worth of data to work with. And there are teams of very smart programmers at Google working to make human interaction with it smoother and more natural.

You might be interested in reading** Accelerando**, by Charles Stross. In it, he theorizes accelerated technological change resulting in the conversion of all matter in the solar system to computer storage, which once reaching critical mass, causes the entire solar system to become sentient. This author and this novel are at the absolute forefront of science fiction’s Big Ideas. Not one to miss, if you’re into this.

I don’t think we know enough about consciousness to rule it out as something that might spontaneously emerge from a sufficiently complex system like the Internet. I’d guess the likelyhood of it happening is vanishingly small, though.

The Bad News: The Internet will soon achieve conscious self-awareness, and super-human intelligence.

The Good News: Almost the only thing on its dirty little mind will be pornography.

Humanity dodges the bullet once again.

So when do we start seeing cyborgs that look just like humans, only they’re out to kill us?

First, ditto what jawdirk and friedo said. Second, having recently written one of those bad sci-fi novels, I did a fair amount of research on this. I’m no computer scientist, but I’m pretty confident that the answer is “no friggin’ way.”

The main difference between mechanical intelligence and biological intelligence is that programs are only capable of analyzing existing data. There is no programmatic mechanism for spontaneous thought, only the simulation of it via randomness. You can’t program a computer to come up with an original idea or impulse. That’s not to say that programs cannot be designed to learn, which they certainly can, and that might be a step toward creating some kind of self-awareness. As far as I know, however, there’s not even a computational model for the kind of cognitive leaps that biological processing is capable of.

It’s possible that we will eventually be able to reduce biological processing to the point that we discover it is ultimately based on randomness as well, which would mean there was no fundamental barrier to creating a program that did the same thing, but that doesn’t seem to be the case so far.

It might be possible to tap the internet as a source of massively parallel processing, a la SETI At Home, to create a calculating machine with the equivalent power of a human mind, but this would have to be a deliberate undertaking and could not happen accidentally. Or we could just wait around till processing power becomes so cheap that we can assemble enough parallel processors in a single machine to simulate the number of impulses a human brain can process.

But sheer computing power is not the same as consciousness. Almost by definition, a conscious machine intelligence will need to have the capability to alter its own programming. Otherwise, it will simply be processing according to a human-designed routine, no matter how complex that routine happens to be. (The internet certainly does not, and will never qualify as “self-programming” – at least not in its current incarnation.)

I think that true “Artificial Intelligence” in the sci-fi sense will eventually become a reality, but not through pure accident or the simple accumulation of processing power. Rather, it will be through the synthesis of research into how biological processing works and deliberate development of computational models to simulate that. It will almost certainly appear in an academic environment first.

Interesting, Scupper.

My take on the future of AI is that it will show up eventually, if only in so much as computers become so good at imitating human functions and interactions that they will be all but indistinguishable from non-artificial intelligence. Indeed, they already have programs that you can type a conversation with that will simulate a human interaction. Of course there are still some flaws to work out, but as they get better and better, they’ll be harder to “stump” and show to be a program. IMHO, the point at which computers “come alive” is the point at which you can’t accurately tell if they’re computers or not.

Think about it, how can you be sure that a human is typing this post right now? Maybe it’s just a cleverly written posting-bot that can accurately simulate message board behavior and self-referential musings.

Well, of course there are the speelling errers.


To be fair, the ability for a node in a connectionist network to be able to “reprogram” itself isn’t strictly needed for the network to exhibit intelligent like behaviour. Either the way the nodes are connected to one another, or the importance placed on the output of a particular node by another node being altered lead to intelligent like behaviour (or both).

Apologies if you already knew this.

No, it’s pronounced “Al-gore”.

Heehee. And yes, I know he didn’t really make the Internet.[

Where is this AI going to be developed? Honestly, I’m an AI student and I can’t see any computer becoming sentient in my lifetime or my children’s and grandchildren’s lifetime. The only way that I can envisage it is if it happens by accident, and even then, what good is it if we don’t understand how or why it works the way it does?

Seems like there was a movie (no, not Demon Seed) based on this premise (pre-Internet, maybe late 60s early 70s) where a US super network hooked itself up to a Sovient super network and became intelligent and decided to take over the world. In the process it imprisoned a man (scientist?). The man finally convinced the Network that he should be allowed a conjugal visit without being observed, and somehow cooked something up with a woman who help him defeat the Network. Can’t think of the name and couldn’t come up with the right search criteria to find it in IMDB.

Try talking to Alan.By no means perfect, but still impressive (to some).

The 1970 movie Colossus: The Forbin Project, perhaps?

Colossus: The Forbin Project. Based on a series of books by D.F. Jones.