At the 10th International Summer School on AI and Big Data, Dr. Rudolf Seising will talk about Fashions of Artificial Intelligence.

Talk: Fashions of Artificial Intelligence

When did the term Artificial Intelligence (AI) come into the world? What did it mean and how has it developed since its emergence in the mid-20th century? AI was coined as a term for a field of research in 1955, when the young mathematician John McCarthy at Dartmouth College in Hanover, New Hampshire, planned a “Summer Research Project on Artificial Intelligence”, which he, together with Claude E. Shannon, Marvin Minsky and Nathaniel Rochester, proposed to the Rockefeller Foundation to fund. Having previously failed with his proposal of “Intelligent Machines” as the title of the volume he and Shannon then published as “Automata Studies”, McCarthy then pushed through the use of the word “intelligence”.

For cybernetics, the “superscience” launched by Norbert Wiener in the last third of the 1940s, the fast-computing digital computers built in the USA at that time became the paradigm of the “rise of the machines”: Machines that can control themselves and learn. Here, a thinking of man from the machine emerged, but also a thinking of the machine from man (Thomas Rid). Cybernetics prepared AI in different ways in different contexts. From England, William Ross Ashby tried to bring his view of the new computers as “amplifiers” of human intelligence into cybernetics; Philipp Aumann attested that cybernetics in Germany was characterised by utopias and speculations, which is why he called it a “fashionable science” (“Modewissenschaft”).

Even in earlier times, speculations were made about whether machines exhibit intelligent behaviour, and Alan M. Turing discussed these in his 1950 article “Computing Machinery and Intelligence”.

The philosopher John Haugeland described AI as a fashion – “Good Old Fashioned Artificial Intelligence” (GOFAI) in his 1985 book Artificial Intelligence: The Very Idea. In this way, 30 years after the Dartmouth event, he distinguished these AI methods of symbolic representations, which have since been called “classical”, from newer approaches that do not use explicit high-level symbols, such as mathematical optimisation, statistical classifiers and neural networks. These “new-fashioned AI”, “new wave AI” or “new-fangled AI” (NFAI) researches almost totally shape our current concept of AI, mostly without considering its historical developments and meanings. In this lecture, I present a view on the history of AI and these mentioned fashions.

Bio

Rudolf Seising obtained his Ph.D. in Philosophy of Science and the Habilitation in History of science from the Ludwig-Maximilians-University (LMU) in Munich after studies of Mathematics, Physics and Philosophy at the Ruhr-University of Bochum. He was acting as Professor for the History of science at the LMU (2009) and at the Friedrich-Schiller-University in Jena (2014-2017). Dr. Seising was Visiting Researcher (2008-2010) and Adjoint Researcher (2010-2014) at the European Centre for Soft Computing in Mieres (Asturias), Spain and he has been several times Visiting Scholar at the University of California, Berkeley. He managed the historical research project “IGGI – Ingenieur-Geist und Geistes-Ingenieure: Eine Geschichte der Künstlichen Intelligenz in der Bundesrepublik Deutschland” (“IGGI – The Engineering Spirit and Engineers of Mind: A History of AI in the Federal Republic of Germany”) at the Research Institute for the History of Technology and Science of the Deutsches Museum (2019-2023).