ABSTRACT
Anthropic artificial intelligence (AI) will be defined as pursuing the computer implementation of the base, minimal, human ability that allows for our culture but is independent of it. Anthropic AI aims at minimally human intelligence – without presupposing any of our cultural heritage. Having defined anthropic AI as the base, minimal, human ability that allows for our culture but is independent of it, we can turn to examining what the alternative approaches to human-like AI may be. Less effort is being invested in human-like AI in comparison to other technologies, and very little effort in teaching this field, the main effort that is ongoing in creating human-like behaviour is chatterbots for commercial and Turing test purposes; this is human-like but not anthropic. Anthropic AI assumes a pragmatic distinction between the layer of intelligence that allows for human learning on the one hand and the content of such learnings on the other hand.
