Synthetic Study on Some Drugs and Bioactive Compounds Inspired by AI System SYNSUP
概要
1.1 Artificial Intelligence
In modern society, computers and networks are common infrastructure. Advanced
software has been developed for various purposes in every industrial field. The first
electronic general-purpose digital computer, ENIAC, was released in 19461. Similar to
many other scientific and technological innovations, it was intended for military purposes
such as the calculation of ballistic trajectory and simulation of nuclear fission. In the
meantime, the famous British mathematician and computer scientist Alan Turing, who
presented the theoretical concept of the stored-program computer2, formulated a chess
program as early as 19483. Once personal computers became available, chess programs
that operated at the amateur level were released. Since then, the algorithm has been
improved as microchip technology has advanced. To display its power, IBM, the computer
giant, challenged the reigning world chess champion, Garry Kasparov, to a match against
their chess-specialized computer Deep Blue. While Kasparov won the first challenge,
Deep Blue succeeded in winning the second match in 19974. However, there was some
controversy regarding to what degree the computer was backed up by human intervention
in terms of the selection of opening moves and suggestions in the middle of the game5. In
recent years some chess programs have been developed that could compete with chess
grandmasters. The evaluation of chess positions has been improved through machine
learning of the records of games and trial matches with other computer chess software, but
it is still beyond the power of machine learning technology to trace all the possible chess
moves from the beginning to the end.
Artificial intelligence (AI) research has been carried out in various fields and yet, the
degree of achievements was much less than previously anticipated, and this brought about
disappointments in its history (Figure 1-16). There have been two so-called “AI winters”
where research activities were diminished due to criticism and financial setbacks. Boom 1
has been labeled as “GOFAI,” which stands for “Good Old-Fashioned AI.” The Samuel
Checkers playing program was among the world's first successful self-learning programs,
and as such, a very early demonstration of the fundamental concept of AI7a. In Boom 2
expert systems were developed, which was an approach to analyze the expertise of
humans and make judgment rules for the selection of candidate solutions. ...