Are you passionate about revolutionizing Human-Computer Interaction? Could you see yourself developing real-time, energy-efficient AI inspired by the human brain? And are you interested in exploring Bayesian models of human behavior for cutting-edge Neuromorphic HCI? If so, apply for this PhD position at the Human Media Lab!

Recent developments in Large Language Models (LLMs) have shown promise for the delivery of user experiences that fulfil the long-standing concept of the digital assistant in Apple’s 1987 Knowledge Navigator video. However, there are many outstanding issues that impede progress towards a seamless interaction with digital assistants. For one, the input of data about the user’s ongoing behaviour is not available to LLMs. This means LLMs can have difficulty reasoning about everyday user activities. Secondly, available data is not used to train the model in real-time. Thirdly, current AI does not explicitly understand the uncertainty that underlies human behaviour. This means LLMs have difficulty reasoning about the reliability of their inferences, which can lead to difficulties in real-time decision making. Finally, training LLMs is costly from both a computational and environmental perspective. 

Requirements 

• You hold or will soon hold a Master’s degree in Human-Computer Interaction, AI/Computational Neuroscience, Computer Science, Electrical Engineering or a related field from a top school.
• You have excellent empirical, programming, maker or electrical engineering skills.
• You have an interdisciplinary mindset, with a strong interest in AI for human factors as well as programming/engineering/making skills.
• You are proficient in English and have good communication, presentation and writing skills in English.