Characterising Artificial Intelligence technology for international transfer

Abstract

One of the central factors influencing the process and the outcome of technology transfer is the nature of the technology being transferred. This paper identifies and discusses the main characteristics of Artificial Intelligence (AI) technology from the point of view of international technology transfer. It attempts to indicate the peculiarities of AI in this context and move towards a framework to assist recipient decision makers in optimising the formulation of their policies on AI technology transfer.

The five AI characteristics identified here relate to complexity, localisation, uncertainty, capital intensiveness and awareness. Some of these features are in principle common with other high technologies, but still bear some aspects specific to AI, albeit only the intensity with which they characterise the technology. The complexity of AI technology partially stems from its multi-disciplinary nature. Certain sub-areas of AI are locally bound and therefore need local innovative capabilities to develop on the fundamental concepts. The uncertainty in AI projects stems mainly from the two factors of high rate of change and the difficulties of quantifying the gains of transferring cognitive load from human to machine. Expensive human skill and enabling technologies are required to develop, maintain and use AI systems efficiently, making such projects highly capital intensive. Finally, AI is a young field and is not always completely open about itself which makes the decision makers' awareness an important issue to be addressed in the growth of AI applications.