New Page 1

PREVIOUS PAGE |_Page index_| Website | Review Page | Journey | Donate | Links | NEXT PAGE

 If you jumped to this page from another page from within the eBook, then click your browsers back button to return to it, rather than the previous or next buttons above.

You are reading a page from a free video eBook called Heaven or Hell It's Your Choice, for more information click on the website button above.

Section 4 / Page 143

The Basic Model Continued

If you start from basic principals, i.e. here is a picture (virtual model), followed by the sound and visual characters we use to represent that data, then it should be possible to teach an A.I. system to build up a basic model. If you have kids then you will know this learning procedure, as in here is a picture of a cat, followed by the sound and the visual patterns used  to represent a cat (i.e. the written language).

So if a 3D virtual model of a cat was produced, then an A.I. system could be given the ability to study all aspects of that model, it should then be able to classify that data. As this data structure was in fact based on a real object or in this case a cat, it should then in the future, be capable of identifying that object. The building up of a visual database of both static and moving virtual models, should have allowed the MNN to eventually build up a basic model smart enough to intelligently understand the real world. So just like humans in the old days before we evolved a spoken language, we could draw a cat and then most other humans could recognise it.

Humans then developed language, in this case the ability to say cat and then later on the ability to draw a symbol representing a cat. It's only by combining these 3 elements visual, audio and the symbols we have developed to represent things, that I see a way of producing an A.I. system capable of understanding language, at least in the same way we do. I do not underestimate the software engineering challenges involved in doing this, but I believe itís just a question of time. The concept of the macroscopic neural network and the database network / grid, was to give an A.I. system both the data it would need and the processing power, to comprehend and manipulate all that data in real-time. This is what I believe will be needed to produce the level of A.I. envisioned, at least using current technologies.

The theorists who believe that consciousness is ultimately a function of some genetically inherited or semi hardwired emotional mechanisms built into the brain, maybe right. These are the people (the mysterians) who usually believe that machines will never be capable of self-awareness, but there is no telling how silicon life will evolve. The ultimate complexity of some of the systems currently being dreamed up, may eventually outstrip our own biological complexity, (see Biophysics). In the future, nano fabrication will mean that these things will be built to operate at similar scales (size wise) to biological systems, so they may evolve their own emotional awareness, especially if they are designed to mimic our own neural matrixes. Keep in mind, that machines may not need emotional intelligence so as to have a very high IQ. I say this because there have been studies done on some humans, who exhibit very low EQ and yet have a very high IQ.

Anyway, I find self-awareness rather than intelligence a fascinating subject in this context. Is awareness a function of some bio or electrochemical stimulation within certain parts of the brain or is it purely a function of intelligence, itís hard to know for sure, but I am sure, that external stimulation from our sensors is a major trigger involved in the development of self-awareness. These triggering mechanisms are a basic function hardwired into our brains, in other words, the fact that our brains are hard wired into our nervous system, also means that our brains have a constant input of stimuli, itís this stimuli, which goes to configure our own neural matrixes, so I would like to use this simple analogy:-

The stimuli inputted into the brain is just a set of complex electrical impulses whizzing around a circuit, (neural pathways), so allowing those impulses to be processed. This is my reasoning for arguing that computers can become self-aware, i.e. if you build a brain or in this case a large grid setup, that incorporates increasingly sophisticated transivity mechanisms, then it should eventually receive enough stimuli (input) and in real-time, that it will eventually evolve, into something much more capable and greater than the sum of its parts, just like us. Humans have managed to evolve some of the most complex feedback systems this planet has ever seen, but this may pale into comparison, to future A.I. systems?

The building up of a computerís possible self awareness or basic model, in the way just described, is going to be a tough job, but I do feel that it is the most probable and best available route currently open to anybody wishing to develop truly smart or even possibly self aware A.I.. The MNN's basic model was not envisioned as being a piece of hardware, but a piece of evolved software, which was universally capable of being added too, somewhat like our own minds, although the dangers in doing so, are highlighted throughout this section of the eBook. Once the basic model had reached a certain level, then its ability to learn should have been controlled, otherwise the problem of having something smarter than humans roaming around could arise. The science fiction rules of robotics may have to be introduced, that's if mankind does not want to feel like an inferior species next to these artilects (Artificial Intellect) as they have been called. (See Isaac Asimov). I envisioned many different basic models being produced and then applied to do specific tasks. Just like a set of macros, that grow in complexity, just as they do in today's software designs or in laymen's terms, many repetitive clicks are eventually learnt and then transformed into a single click operation.

I think, Asimov's rules are currently impossible to implement on today's A.I. and robotic systems, but maybe in the future, when these things have grown up, it maybe possible, but I personally don't see how it could be done.

The Basic model described in a slightly different way or is it the same way?:-

Please report any problems you see on this page -such as broken links, non playing video clips, spelling or grammatical errors etc to:-

I don't have the time to individually respond to every email, so I thank you in advance, for your help.


PREVIOUS PAGE |_Page index_| Website | Review Page | Journey | Donate | Links | NEXT PAGE

Author Alan Keeling ©