Le Monde | • Updated | By
The Turing test, which is to see if a machine can impersonate a credible way for a human? “A joke! “ In a video interview posted in 2013, Marvin Minsky, himself Turing Award (1969), described as a joke what often passes for the final test that would qualify as an artifact created by smart the man. That was not the intention of Alan Turing, was pursuing the American researcher, in this blog dedicated to the “singularity” that point where machines will definitely sway over Homo sapiens . Intelligence, artificial or not, was not to be locked into a single test, Minsky suggested, having spent most of his career to try to capture the essence for transmission to PLCs – and see that she hid in part to his efforts.
Marvin Minsky, born in New York August 9, 1927, died of a cerebral hemorrhage, Sunday, January 24 in Boston (Massachusetts). It is considered one of the fathers of artificial intelligence, but its scientific eclecticism also led him to invent the confocal microscope and to propose, with others, a concept of spatial fountain that would offer permanent access to space.
At the end of the Second World War, where he served in the Navy, he began studying mathematics at Harvard University and then at Princeton, where he defended his thesis. Those years of apprenticeship, he recounted, put him in contact with many researchers who fled Europe during the rise of Nazism and war, he could pass on campus. – Like Albert Einstein or Kurt Gödel
“Designing a new discipline”
Besides his childhood readings of Jules Verne, HG Wells or later Asimov, who were to influence his choice of a scientific career – he was also excellent pianist – Minsky quoted from the works that have marked the Mathematical Biophysics to Rashevsky Nicolas (1938). This text was put on the mathematical modeling of biological phenomena route, including how neurons can learn. In 1951, he proposes the stochastic analog neural reinforcement calculator (SNARC), the first neural network simulator. To better understand the operation of real neurons, he began to develop a new type of microscope, said confocal (1953), including improved versions are still used in biomedical laboratories.
In 1956, it is one of the organizers of a summer school at Dartmouth College (Hanover, New Hampshire), with Claude Shannon (the father of information theory) Nathaniel Rochester (IBM Scientific Director) and John McCarthy, young researcher credited with coining the term “artificial intelligence”. “Their idea was to design a new discipline, and Minsky had a major contribution” said Jean-Gabriel Ganascia the computer lab of Paris-6.
In 1959 he founded with McCarthy Artificial Intelligence Laboratory at the Massachusetts Institute of Technology (MIT). With Seymour Papert, a childhood friend, he developed performance analysis tools artificial neural networks. In their joint book, Perceptrons (1969), “It shows the limits of machine learning, which will put a temporary stop to the research in this area” notes Jean-Gabriel Ganascia. His responsibility in this “first winter” artificial intelligence will debate. Minsky then launches into the study of semantic memory
.
big fan of science fiction
Many of his books will mark generations of researchers. He was particularly pleased with The Society of Mind (1985), describing intelligence as the result of cooperation of various agents, and could be read without having a significant scientific background. He feared, however have lost its readers with The Emotion Machine (2006) more difficult.
He regretted the liberal turn taken in the 1980s by the United States, that would have slowed the drying up basic research in such financing on artificial intelligence by . the Office of Naval Research, loyal support of the discipline since the 1940
It was critical to the recent mega-brain modeling, like the Human Brain Project – which he believed ” 98% “ chance he does not allow to understand how the brain works -, fearing they will penalize research in this area
A great lover of Science magazine. -Fiction Astounding Stories , he told Asimov chose not to visit his laboratory where he worked in the development of robots, not dry up his imagination at the sight of poor performance of its machines . It would have advised Stanley Kubrick during the preparation of 2001: A Space Odyssey . In recent years, he had approached the themes of the singularity, and had supported a company of cryonics. Regret to some observers, who saw it as a manifestation of his excessive taste for science fiction
Read also:. The surpromesses artificial intelligence
No comments:
Post a Comment