AI Artificial Intelligence – and the story behind it. We start a small series with highlights from several decades of AI development and begin in 1945: In his 1945 article “As We May Think,” Vannevar Bush outlines his thoughts on the storage of increasingly specialized knowledge against the backdrop of the rapid technological progress of the time. Starting from the common index management of libraries, he thinks about an association-based management of knowledge. In his mind, books should be stored in miniature and retrieved by means of standardized codes. The innovation, according to Bush, is to be able to view several books simultaneously and link them to each other. Today, this would be called a (graph-) database.
Using this process, he describes the creation of a new book that contains elements from all the linked works. From today’s point of view, he has described something that can now be found as a matter of course in every PDF file. Especially scientific (review) articles benefit from this early conceived feature. The postulated machine has the name “memex” , it is said to have several output screens, a keyboard and a “set of buttons and levers”. In terms of storage capacity, memex is said to be capable of not reaching its capacity limits for hundreds of years with a daily input of 5000 pages.
Vannevar Bush’s thoughts outline the understanding of Artificial Intelligence (AI) at the time: an association-based database. If we think of the autocomplete function in every Google search or the suggestions from Amazon, the AI of those days has long since been integrated into our everyday lives.