SYMBIOSIS is our mission to understand and co-develop human and artificial intelligence in real-world settings.

The pillars that underpin SYMBIOSIS are:

Multi-scale models of intelligence as a closed loop

  • Using multi-scale models, we study intelligence as a closed loop, observing humans and AI as interconnected systems whose components continuously interact under real-world constraints.

Nature-efficient AI

  • Can AI reach the efficiency of biological intelligence in the 2030s? Our goal is to create nature-inspired AI hardware and algorithms, guided by lessons from biological systems, including the human brain.

Embodied intelligence

  • Modelling how thought, movement and emotion are integrated as one to shape intelligent behaviour in humans - and applying these insights to guide the design of new AI.

Physical world AI

  • Advancing how AI systems can move, make decisions and work alongside people in everyday environments. This includes training machines to navigate spaces, respond to change and act safely in human-centred settings.

The empathic machine

  • Can we augment machine intelligence to interact with human thoughts and emotions in an ethical, unbiased and culturally aware way by the 2040s - so that people won’t just use machines, but feel understood by them? 

A virtuous cycle of understanding and change

  • We use living laboratories to study intelligence in real-world settings, creating a virtuous cycle of testing and feedback that advances human-AI understanding, inspiring technologies that will have a tangible benefit to society. 

Through SYMBIOSIS, we aim to generate insights from how humans and AI operate together, guiding systems that are safe, adaptive and robust.