AI researchers are - like the alchemists of old who sought to create gold from base metal - seeking to create thinking machines from infinitesimally small bits of silicon oxide.
The birth of AI was tied to the efforts of a variety of talented, intellectually self-confident, well-educated budding mathematicians, electrical engineers, psychologists, and even one political scientist. The various origins of the creators of AI and the enormous influence of their work explain some of the colourful aspect of their pronouncements, and the rollercoaster evolution of their field. As the hoopla over cold fusion illustrated in 1989, even the staid science of physics is not immune to exaggeration and false claims. Yet excesses of optimism seem to occur with particular frequency in AI.
First, there were plausible reasons in AI's early years for believing in its rapid progress, and these induced early researchers to display the excessive optimism that came to characterize the field. Early progresses in using computers for arithmetic were truly breathtaking. In a few years, technology went from cranky mechanical calculators to machines that could perform thousands of operations per second. It was thus not unreasonable to expect similar progress in using computers as manipulators of symbols to imitate human reasoning.
One misconception further enhanced this temptation. ...
 There is truth in this statement, but bridging the gap between ant and human still requires a giant leap in complexity. Yet, in the post-war decades, controlled studies on our ability to remember and reason showed their basic limitations.
The next time you look up a seven-digit phone number, try thinking of something else for a few seconds before dialling. If you are like most people, this will make you forget the number. We cannot keep more than five to nine items at a time in our short-term memory; and as soon as we look away, they vanish. Our long-term memory has an almost unlimited capacity, but it learns very slowly. Transferring an item from short-term to long-term memory takes several seconds. When we rate alternatives in any complicated problem, like a number puzzle, we need pencil and paper to make up for these deficiencies of our memory.
Early AI researchers reasoned that their computers did not suffer from these limitations. Even in those days, the machines had memories with capacities of thousands of items and access times of microseconds. They could shuffle data much faster than humans can. Computers; thought Simon and his colleagues, should be able to take advantage of these capabilities to overtake humans: it was only a matter of a few years before suitable programming would let them do it. These researchers had not realized that, in activities other than purely logical thought, our minds function much faster than any computer yet devised. They are so fast, in fact, that we are not even conscious of their work. Pattern recognition and association make up the core of our thought. These activities involve millions