A recurring question in AI today goes back to an article by Alan Turing in 1950, which reads, "I propose to consider the question: 'Can machines think?'" (Computing Machinery and Intelligence. Mind, Vol. 59, No. 236 (Oct., 1950), pp. 433-460; in the original: "I propose to consider the question, 'Can machines think?'").

In his keynote address "Cognitive IT: The Dawn of the Age of 'Machine Thinking'" at COGNITIVE 2019 on 8 May 2019, Prof. Ferscha reviewed the evolution of early AI, assessed the current state of achievements, and ended in a framework for "embedded AI" in which embedded, wirelessly connected tiny computing platforms with a multitude of miniaturised sensors collect data about real-world phenomena, analyse and interpret this data in real time, reason about the detected concepts and symbols, make decisions, and influence or control their environment via a multitude of actuators. Sensing, reasoning and controlling thus closely connect the physical and digital domains of the world, with feedback loops connecting one domain to the other. This is how they implement concepts of autonomous, adaptive behaviour.