Artificial Intelligence: A Century-Old Story

Abstract

Artificial intelligence (AI) origins can be dated all the way back to 1948. Alan Turing wrote the first AI manifesto called “Intelligent Machinery,” theorizing that in order to be considered intelligent, a machine should be able to imitate human behavior so as to be indistinguishable from human themselves. In the same period, McCulloch and Pitts first introduced a prototype of neural networks. It was in 1955 at the Dartmouth Conference that the term “artificial intelligence” was used to describe the science aimed to develop machines that simulate human’s behavior. From that point, research in the field of AI never stopped, being a rollercoaster of successes and failures. In more recent times, thanks to the development of more powerful computers and efficient algorithms, the field of AI and especially that of neural networks are gaining more and more attention, reaching some remarkable results such as the defeat of Garry Kasparov in a chess match and the ability to autonomously drive a car. AI is being implemented in multiple areas, including medicine and more importantly medical imaging. Throughout the years, software that can support physicians in making diagnosis, select optimal therapy options, and assess prognosis have been developed. In this chapter, we are discussing some of the crucial steps that led AI development, giving insight in the major hurdles AI has already overcome and which ones are still present in the current age.KeywordsArtificial intelligenceBig dataComputing powerDeep learningRadiomics