Citaten
At best, AI is engineering around the problem by simulating intelligence, rather than facing the problem by inducing intelligence.
A known Chinese saying applies to Semantics: "Give a man a fish and he will eat for a day. Teach the man how to fish and he will be fed for a lifetime".
I am developing an information system that knows how to "fish" - creating its semantics autonomously from the logic contained within grammar - rather than feeding such a system a semantic vocabulary and subjects of personal interest (everyday).
Evolutionary algorithms are purposefully designed to operate within specified boundaries in order to reach a specified goal.
Evolutionist Richard Dawkins - author of The Bilnd Watchmaker - would kick your butt, if you use words like purpose, design, specified, goal, or boundary in the same sentence as the word evolution.
Illustrating the scientific approach to AI by a metaphor - the dream of flight:
Many have tried to replicate birds to fulfill the dream of flight:
• Some were sure the magic of flight is in the bird's feathers;
• Some were sure that the complexity of the bird's wing construction would reveal the secrets of flight;
• Some were sure that just a more powerful engine would make their dream come alive;
• Some were sure that humans could never fly, because they are just too heavy;
• However, the Wright Brothers experimented for years to understand the fundamentals of (the physics behind) air lifting.
History repeats itself in AI:
• Some are sure the magic of intelligence is in connecting neurons to ANNs (the feathers), in the hope intelligence will appear by "itself", whatever that is;
• Some are sure that intelligence is in complexity, but they forget all the quotes from Albert Einstein and others that state: intelligence reflects in simple solutions;
• Some are sure that just more CPU power and/or larger databases would make their dream come alive: Current hype: Social media will deliver better semantics: more text = more meaning. Really?;
• Some are sure that AI will never happen;
• Compared to the metaphor, simulating intelligent behavior is like building a flight simulator - rather than building the real thing: a flying airplane.
Complexity doesn't occur by increasing quantity, but by lack of quality and understanding:
Complexity only occurs when lacking structure, management, laws, rules, procedures, logic, overview, etc. Take all those disturbing factors out of the equation and your problem is not complex anymore.
An example: According to IBM, Watson is able to find the needle in a haystack of unstructured texts. But all they did is: engineering around the problem of intelligence, because the real problem is: unstructured texts.
I however, face the problem of intelligence, by developing a system capable of structuring texts autonomously, taking the disturbing factor of "unstructured texts" out of the equation, which makes a search more like browsing a library for a topic.
We really should distinguish the use of the normal word evolution - meaning: changing or developing - from the polluted meaning referring to the evolution theory:
Evolutionary computation techniques are algorithms, purposefully designed, allowing variables to change within given boundaries.
However, according to evolutionist Richard Dawkins, evolution - as assumed by the evolution theory - has no design, has no purpose, no boundaries, etc. So, evolutionary computation techniques have nothing to do with that magical power of evolution assumed by the evolution theory.
And therefore, we should be very specific when we use the word evolution.
If fact, evolutionary computation techniques actually invalidate the evolution theory in an indirect way, because they are designed purposefully and operate within given boundaries. Three words that are in conflict with the evolution theory.