So Is The Web Exhibiting Sentience Now?

I’m with Evil Captor. Looking at a single neuron gives no clue that hooking up enough in the right arrangement results in consciousness.

What if “God” is the conciousness that will be created when enough of us are connected by the Internet?"

Are you familiar with the old story that some dignitary asks a supercomputer “Is there a God?” and the computer responds with “There is now.” ?

For those who think it is possible, do you know how computers and the Internet work in intimate detail?

If it is possible I think it would require orders of magnitude more bandwidth and processing power than we have now.

For those that think the internet or google or something like that will gain sentience merely through shear volume of data and computation, consider the following:

  1. If you took 10 billion neurons (and other cells) and threw them all in a bucket - would you have something anywhere close to a working brain? (hint: no)

  2. Why did evolution fashion our brain into just the right structure to achieve sentience? Because it helped us survive/procreate. What would guide the internet/google to sentience? What are the pressures causing tweaks and changes that would move it in this direction? (hint: nothing).

I mostly, but not entirely, agree with you here… I think that search engines, expert systems, and other problem solving algorithms could – in the abstract – begin to operate on themselves, making themselves more effective, with some sort of sentience as a result. I don’t think that’s how it will actually happen, and certainly not happening now… But a general “problem solving” system, if applied to itself, could improve itself into sentience… I think…

(therefore I eat spam…)

I think it’s what this book says: The Global Brain

The reader reviews will give you a good idea of what’s in the book.

Right now the Internet consists of 2 aspects - data and web applications which are isolated from each other. Data by itself does nothing, while the various web applications do not communicate with each other in any meaningful way that resemble sentience. Those applications are specially tailored for a purpose.

An AI is a very deliberate piece of programming, and different type of problems require different programming algorithm to solve. The most generic one I can think of is neural network, but like all programs, it’s garbage in, garbage out. Until we have cheap ways (in terms of computing power and speed) to analyse data and place them in their proper context, AI the like of EDI and HAL are still far off. Even so, an AI program is usually tailored to a single purpose - example, getting from point A to point B. Within those specialized program, as data accumulate, the AI may act in unpredictable manners, but an AI which only knows how to go from one point to another won’t “figure out” how to use a gun, for example. It’s not in its “programming” to do so.

Likewise, if a piece of AI program, which is able to mine the Internet for data, will only be able to work with data it is taught to recognize and do stuff with them as it is programmed. The final decisions of the AI may be out of whacks over time, as it learn new rules and such, but I don’t think at the moment an AI program can go out of it boundaries. For example, Google search can crawl website and retrieve results, but it is unlikely able to start chatting with you all of a sudden by itself unless the Google programmers add a chat module and the AI for processing speech.