Article:

Menschengestützte Künstliche Intelligenz. Über die soziotechnischen Voraussetzungen von «deep learning»

The current success of artificial intelligence is based not only on technological advances, but also on a fundamental sociotechnological structural change. Machine-learning methods such as deep learning require large amounts of training data that can only be obtained through human collaboration. In a convergence of methods of human-computer interaction and AI, a wealth of human-machine interfaces and media infrastructures has emerged over the past ten years, through which human cognitive resources are harnessed in hybrid human-machine apparatuses. These devices perform, as a whole, what is known as artificial intelligence. AI in its present form is thus a distributed, networked phenomenon.