How computers came to know the world in two ways – making computer models in technical academia in Denmark.

Description

In January 2022, the Danish magazine Ingeniøren professor of social aspects of artificial intelligence Thomas Bolander took stock of AI. He identified two individual paths. The first path is based on language, rules and logic. The other path is machine learning with statistical models and learning through examples. According to Bolander, these two paths are disconnected and have developed as individual paradigms. In this paper, based on studies of technical academia and related companies in Denmark, I will discuss how these two paths developed, the risks in using these models and if these risks are familiar to us. The first path is the oldest. A link between science and engineering was cultivated in technical academia and mathematical scientific theory based on equations was used in predictive practices and, in turn, when making decisions. Technical academia and engineers would strive to create new measurables, define parameters and set standards to make stronger predictions. These were the basis of engineering models and technology-based decisions. The second path opened up later as it is dependent on advanced machinery. In machine, learning the computer is trained through examples and experiences, while it looks for patterns. In technical
academia, these practices are connected to example based teaching. In engineering training students were shown a wide range of pictures or studied samples in large numbers to train the feel for the good solution. Students also took part in numerous drawing exercises to train what Ferguson calls the mind's eye.

Period	24 Sep 2022 → 16 Oct 2022
Event title	Conference International Committee for the History of Technology: Technology-based and Technology-generated Decicions
Event type	Conference
Conference number	49
Location	Ostrava, Czech RepublicShow on map
Degree of Recognition	International