We're all doomed!!!!!

[GreyGeek]

I actually think it's right around the corner. Check out some of the stuff http://numenta.com/ is doing. They have actually developed software that is emulating at least some of what the neo cortex does, and are using it in commercial applications. It's pretty neat stuff and it's not "traditional" AI.

Also keep in mind that these kind of advances tend to happen exponentially. E.g once they emulate the brain of comparable to say a mouse, the human brain is not far off.

The web site you link to is exactly what I am talking about:

What Is Machine Intelligence?

Because today’s computers are programmed, they can only do exactly as they are told. In stark contrast, intelligent machines continuously and automatically learn patterns in their environment without being programmed, enabling them to tackle problems in entirely new ways. Intelligent machines that learn and act will have an enormous beneficial impact in the coming decades.

First, that is PR hype designed to sell software.

Secondly, "Machine intelligence" does exactly what it is told to do as well. It IS running software, not flinging pixie dust around. And, contrary to popular opinion, it is not rewriting that software as it goes. Consider "learns patterns in their environment". How does it "know" what a pattern is? By comparing it to patterns it has been programmed to recognize as it built up its stock of neural network pattern solutions. These systems can get very good at recognizing patterns of numbers in hand written scribbles, for example, or in recognizing packet patterns or electrical fluctuation in servers, but that isn't "thinking" or even intelligence. That the Numata's first product, Grok, can handle computer server performance but could not transition by itself to facial recognition or guiding a vehicle is proof of its limitations.

The Numata builds on work by O'Reilly and Rudy in papers written in 2000 and 2001:

Our framework accommodates this finding by establishing a principled division of labor between the cortex and hippocampus, where the cortex is responsible for slow learning that integrates over multiple experiences to extract generalities, while the hippocampus performs rapid learning of the arbitrary contents of individual experiences. This framework shows that nonlinear discrimination problems are not good tests of hippocampal function, and suggests that tasks involving rapid, incidental conjunctive learning are better. We implement this framework in a computational neural network model, and show that it can account for a wide range of data in animal learning
,...
Additional principles concern the nature of learning (error-driven and Hebbian), and recall of information via pattern completion.

That's the big print. The fine print is the code they wrote to enable neural network pattern recognition combined with sophisticated (but not intelligent) data collection, with input and out code to handle the data. No "thinking" is taking place.

I've been paid to write neural network models of the commodity market and studied them extensively, including the calculus behind them. When I read about the claims of AI it is easy to see all the mental hand waving proponents do to support their contentions, relying on the neophyte or unlearned to assume in their mental evaluation "facts" that are not in evidence.

Claiming that they are modeling cortical learning techniques Numata introduced a programming language called NuPIC, which impliments a "Cortical Learning Algorithm" (CLA). They even supply a NuPIC Studio to provides an HTM construction toolkit and 3D visualizations. For those not knowledgable about programming the term "studio" is used to describe programming tools which encompass a complete method for writing specific software. Using their software development GUI you can simulate their approach. Notice that while they change the naming terminology to neurological terms the approach is essential the old multi layered approach with weighted feedbacks and feed forwards with non-linear detection methods to determine boundary conditions. As you use their tool and change your camera position, identical to changing the camera position in a POV model of a house, you can see where the intelligence lies: in the brain of the programmer running the software and setting up the conditions and properties of the network.

Stanford is teaching a class in writing software using the Neocortex model and gives many examples:
http://web.stanford.edu/class/cs379c/readings.html

[woodsmoke]

I, personally, think that this kind of opinion is a dependency of the philosophy that "humanity" is a "zero sum game".

There were, and are, others, such as R. Buckminster Fuller, who think that "humanity" is not a "zero sum game" and, just as in the past, humanity has always somehow "engineered" our way out of situations, we will so do in the future, given that "politics" does not become so egregious as to bring about the downfall of humanity.

So for me, it is "parts of humanity" which can bring about the demise.

An example from a few years ago was that USA Aid sent a shipload of food to a country that had a monsoon and the food sat on the docks while the king put a label over the USA aid label that had his name on it (can't find a linky right now). In other words, one guy caused the problem, not "humanity".

The internet is always ablather with horror stories of "this and that", but.........the REAL point of that is somehow missed. The internet is ablather. That means that the problems are being exposed. And, there are two reactions.

a) the reaction of "the internet blogger talking heads" which is "usually" glass if half empty, and zero sum game, and the world is coming to an end.

b) the other reaction is the opposite, that somebody takes action, thus illustrating the glass half full and that things are not a zero sum game.

But.............the second situation is not ballyhooed, it is not talked up, it is not even discussed because, the blogger talking heads have run to the next emergency.

Just a thought, of little worth.

http://en.wikipedia.org/wiki/Buckminster_Fuller

Great nations are simply the operating fronts of behind-the-scenes, vastly ambitious individuals who had become so effectively powerful because of their ability to remain invisible while operating behind the national scenery.

http://www.brainyquote.com/quotes/au...er_fuller.html

woodsmoke