The rapid growth of AI has the potential to aid decarbonisation and help us reach net zero through optimising resources and enabling the energy transition. However, the uptake of AI also comes with a rapidly growing carbon footprint. Training large models, powering data centres, and producing specialised hardware all demand significant energy and material resources. As the world pushes toward net zero, the question is clear: how do we make AI sustainable?

Aims

This webinar will explore practical pathways to decarbonising the AI supply chain. We bring together experts to discuss sustainable AI themes such as:

  • A roadmap to reduce AI waste and improve efficiency: towards sustainable AI ecosystems – Andrea Soltoggio, Associate Professor in Artificial Intelligence, Loughborough University

  • Literal Labs approach: Towards energy-efficient and explainable AI via logic and expediency in learning – Prof Alex Yakovlev, Newcastle University and Literal Lab

  • Nature Computes Better opportunity space, adding a fresh perspective to the discussion on building sustainable AI ecosystems – David Stringer, Programme Specialist Lead from ARIA

Together, these perspectives will map out how industry, researchers, and policymakers can collaborate to build a future of truly sustainable AI — from code to compute.

Recording

ADViCE Webinar Series - Building Sustainable AI

Agenda

  • 12:30 - Welcome and Introduction: An overview of ADViCE and the challenge we’re addressing
    • Beth Medlik, Principal for Clean Energy and Decarbonisation, Digital Catapult
  • 12:40 – A roadmap to reduce AI waste and improve efficiency: towards sustainable AI ecosystems
    • Andrea Soltoggio, Associate Professor in Artificial Intelligence, Loughborough University
  • 12:50 – Literal Labs approach: Towards energy-efficient and explainable AI via logic and expediency in learning
    • Prof Alex Yakovlev, Newcastle University and Literal Lab
  • 13:00 - Nature Computes Better: Sustainable AI Hardware
    • David Stringer, Programme Specialist Lead from ARIA
  • 13:10 – Interactive Q&A: Open discussion and audience questions
    • 13:30 – End

Updated: