Mikihito TANAKA, Faculty of Political Science and Economics, Japan
Alain POTTAGE, Sciences Po Paris, France
Marc SCHOENAUER, INRIA, France
Ahmet SUERDEM, Istanbul Bilgi University, Turkey
Philippe FAUQUET-ALEKHINE, INTRA robotics, Avoine, France / SEBE-Lab de la London School of Economics & Political Science, France
Communicating the science of Artificial Intelligence is challenging. Agreement on a definition of AI has remained elusive, and developments have seen successive false dawns and subsequent AI winters. The science fiction that continuously runs ahead of AI inescapably frames expectations, so that AI’s capacity to overexcite as well as to terrify impedes sober communication. Reflexivity is essential. First, AI itself is increasingly shaping the message and the audience in all forms of communication. Second, scientists routinely recruit AI collaborators to speed up data gathering and analysis. Today, artificial decision-making systems surround us – some embodied as robots, others imperceptible in digital worlds.
Our roundtable presents five different ways of thinking about AI: how the technology is imagined; how we can engage with AI agents as legal entities; how we experience physical robots amongst us; our interactions with AI bots and finally how the media represents AI. The co-existence of these different facets of AI, some hidden and some forcibly present, confronts us with the challenge of how to discuss the science of AI when the term elicits such diverse expectations as to what the key issues are.
We all know how it feels when “the computer says ‘No’”. Attempts to make the world more readily interpretable by AI raise difficult questions over where power resides – in the ‘real’ or ‘digital’ worlds, and how these twins are related to each other. And all this without discussing super-intelligent AI, the singularity, or the ethics of autonomous weapons systems. Challenges indeed for science communicators.