(3089/898) We'd damn well better develop Artificial Intelligence, because the natural stuff is in such short supply these days

216) I don't see any way to put a limitation to the degree of intelligence [a machine] could acquire. The only qualification I make, and I can't understand why it is resisted, is that the intelligence that we will develop in machines will always be alien to human intelligence. [...] Consider how different the intelligence of a dolphin is from that of a human being, or how the social intelligence of a computer will be even more different. [...] I don't understand what that reservation takes away from any ambition the AI people might have.

Joseph Weizenbaum
quoted by Daniel Crevier in
AI: The Tumultuous History of the Search
for Artificial Intelligence (1993)

217) [W]e will probably discover in building artificial minds that it pays to design a little differently than Nature. Much as airplanes have wings but do not flap them, intelligent machines will operate on the same principles as their natural equivalents, but will exploit these principles better. Streamlines, robust, and faster, they may well surpass our minds the way airliners do sparrows.

Daniel Crevier
AI: The Tumultuous History of the Search
for Artificial Intelligence (1993)

218)
We achieve more than we know.
We know more than we understand.
We understand more than we can explain.

Claude Bernard
quoted by Daniel Crevier in
AI: The Tumultuous History of the Search
for Artificial Intelligence (1993)

219) To know that one knows what one knows, and to know that one doesn't know what one doesn't know, there lies true wisdom.

Confucius
quoted by Daniel Crevier in
AI: The Tumultuous History of the Search
for Artificial Intelligence (1993)

220) Running a program in a von Neumann [i.e. serial architecture] computer is like moving your household from New York to Los Angeles in the following senseless fashion. First load the TV set in your car, drive it to Los Angeles, and come back. Take the laundry iron, drive to L.A., and come back. Take the coffee pot, drive to L.A, and so on. To enhance matters a bit, computer manufacturers have recently tried flying between cities instead of driving: they have improved data-transfer rates between the parts of the computer. Unfortunately, this amounts to speeding up the circuitry, and soon bumps into the speed-of-light and heat dissipation limitations I mentioned.

An analogue to the obvious solution - using a van to move all items of your household at once - is not possible in a computer. Each bit transferred between processing units and memory requires a separate wire, and there is a limit to how many of these can be crammed into a machine. Over the years, manufacturers have widened the data path from 8 bits at a time to 32, and even to 128 for large machines. A small improvement, this amounts to little more than letting you move both the coffee pot and the laundry iron together.

In terms of my two-city analogy, the solution adopted by our brains is surprising. In consists in moving Los Angeles to New York and mingling the two cities so you don't have to move at all! In my fable, New York plays the role of processing unit, and Los Angeles that of memory. It turns out that the brain does not make any difference between these two functions: each neuron serves as both memory and a processing unit.

Daniel Crevier
AI: The Tumultuous History of the Search
for Artificial Intelligence (1993)

221) [T]he things we do so well that we hardly have to think about them (watching a movie, reading a novel, holding a conversation, playing tennis, and walking down a crowded street) are extremely difficult for computers, while on the other hand, the same computer can be masterful at tasks that stretch our own abilities, such as mathematics, chess, medical diagnoses, and troubleshooting of electronic circuits. It may be that the paradox can be resolved in the following way: Some of the things we do exceptionally well are very complex tasks that require multiple levels of processing but they are also the result of ten of thousands or even millions of years of evolution in which these processes have been refined to a high degree and "hard wired" into our brains. We do not find the act of speaking or recognizing a face to be immensely difficult because all the processing and problem solving takes place below the level of our conscious mind and appears to us as trivially simple. "Difficult things," on the other hand, involve tasks that are evolutionarily more recent and must therefore be carried out in the arena of conscious thought.

F. David Peat
Artificial Intelligence: How Machines Think (1985)

Note: "3089/898" is the designation I've given to the project of posting all my collected quotes, excerpts and ideas (3089 of them) in the remaining days of the Bush administration (of which there were 898 left when I began).

As of today, there are 825 days remaining in the administration of the worst President ever.