In recent months my colleagues and I have been covering News about the latest in Artificial Intelligence. Like so many other things, these achievements are the result of a long road that began in the imagination of novelists and included some ingenious scams. In this post and those that follow, we are going to make a brief history of Artificial Intelligence.
Although these technologies have been with us for quite some time they had not been available to the general consumer. At least not with the level of quality They have services like ChatGPT.
Table of Contents
A Brief History of Artificial Intelligence
No one knows exactly when humans got the idea to artificially create tools that could do the same things. But, the idea appears in both mythology and literature. Of course, there was no talk of microprocessors or software, but of more or less anthropomorphic constructions.s made with clay, wood or even parts of human bodies.
And, of course, there was no shortage of fakes.
Around the year 800, the statue of the God Amon chose the successor of the pharaoh from among the heirs who paraded before it. After taking him by the shoulder, he gave a speech. This technology would be the envy of MIT and OpenAI were it not for the fact that the software that ran it was a priest hidden within it.
In the following centuries Various machines were built that, powered by water or steam, imitated certain behaviors or performed certain tasks typical of humans or animals.. They were known as automata.
Of course, automata cannot be classified as artificial intelligence. According to the classic definition of Marvin Minsky:
Artificial intelligence is the science that tries to get machines to do tasks that would require intelligence if done by humans.
Turing and the Entscheidungs problem
According to historians, the first concrete step towards Artificial Intelligence was achieved thanks to the Entscheidungsproblem. No, I'm not writing down my sneezes. The word was coined by the mathematician David Hilbert in 1928. The problem behind the term is that of knowing if there are mathematical problems that cannot be solved in a systematic way.
Mathematicians call Decisable problems to those that have two possible answers: YES or NO. Two examples would be
Is x a prime number?
Is X*Y equal to Z?
This type of problem can be solved by anyone following a method in the same way that you can prepare a cake with the instructions of a cooking recipe. It's just a matter of time. Hilbert wanted to know if there were decidable problems that did not admit the use of a method.
Today the name of Alan Turing is universally known, basically for his contribution to deciphering Enigma, the code machine of Nazi Germany, and his conviction for homosexuality a decade later. Few know that one of his first achievements was to answer the Entscheidungs problem.
His method was to imagine the existence of a machine that would solve problems following instructions expressed in mathematical form. In other words, the device read symbols contained on a tape and manipulated them following a series of instructions.
According to Turing, this machine had an unlimited memory capacity thanks to an infinite tape divided into squares, in each of which a symbol could be printed. At all times the machine must have a symbol loaded that it can modify. The modification to which it is subjected is partly determined by the symbol itself. Symbols elsewhere on the tape do not affect its behavior unless the tape is moved and another symbol takes the place of the read symbol.
The components of the machine are then:
- A tape divided into boxs that contain symbols (There is a special symbol for the blank boxes
- a head which reads and writes symbols and moves the tape.
- A record that indicates the status of the machine at all times.
- a series of instructions which from the symbol and status register indicates what should be done next.
Be the first to comment