Today, AI systems and products are becoming widespread in many aspects of everyday life, from insurance to policing and online platforms. However, fully data-driven systems raise essential concerns on their ethics and accountability: can we ensure that machine learning systems do not propagate existing biases and stereotypes from data? More in general, can we ensure that an algorithm is fair to everyone? While fairness has now become an essential concept in machine learning, its adoption still requires a fundamental shift in how we view and perceive intelligent systems.
On May 30th we will host a discussion on this topic with Silvia Chiappa, Senior Research Scientist at DeepMind, who will describe the significance of the problem and some of the work she conducted over the last years.