When AI Goes Awry: Professor Des Higham

The next ML group seminar we are excited to have Prof. Des Higham from the University of Edinburgh joining us for an in-person talk (will be accessible over Google Meet as well).

Title: When AI Goes Awry

Abstract:
Over the last decade, adversarial attack algorithms have revealed instabilities in deep learning tools. These algorithms raise issues regarding safety, reliability and interpretability in artificial intelligence (AI); especially in high risk settings. From a practical perspective, there has been a war of escalation between those developing attack and defence strategies. At a more theoretical level, researchers have also studied bigger picture questions concerning the existence and computability of successful attacks.  I will present examples of attack algorithms on convolutional neural networks, transformer models and large language models, including a new approach that exploits techniques from generative AI. I will also outline recent results on the overarching question of whether, under reasonable assumptions, it is inevitable that AI tools will be vulnerable to attack