Feature PANEL: AI for Sustainability and Sustainable AI

Introduction
The rise of Artificial Intelligence (AI) is radically transforming an increasing variety of sectors. AI can contribute to achieving sustainable cities, to attaining an affordable and clean energy, to creating a circular economy, to improving the quality of education, to enabling a number of different targets by supporting the provision of food, health, and utilities to the population. However, if AI can act as a catalyst for achieving Sustainable Development Goals (SDG), it can generate as well negative effects (Vinuesa et al. 2020). For instance, AI can easily amplify social biases by introducing discriminatory and unfair behaviours, as it has been already proved in areas such as human recruitment and loan assessment (O’Neil 2016). It may also increase inequalities to the extent that it can lead to additional qualification requirements for any job, consequently widening the inherent disparities. An effect of this kind can show up both at a global and local level. We need therefore a Sustainable AI as well, i.e. an AI that can be trusted by the citizens insofar as it complies with fair, accountable, and transparent criteria, as a recent document from China Minister of Science and Technology has emphasised (China National Governance Committee for the New Generation Artificial Intelligence 2019). A number of basic questions will be debated during the panel such as:
‍
• AI, for what?
• Machine Learning, to whom?
• Self-driving cars, to where?

All these questions, even though they can appear quite generic, are intrinsically design questions. Looking across the technological box will be as much important as looking into the box, if designers wish to avoid an essential detachment between human and artificial intelligence (Ananny and Crawford 2016). The challenge ahead AI will not only be a technological challenge “per se”; it will be the governance of the transformation processes involved by the use of AI technologies (Crawford and Calo 2016). This is eminently a design task (Floridi 2019; Fabrocini and Terzidis 2021).

• Time & Date: 1:30-3:00pm, 6th December 2021