The lively panel features:
Jono Evans (Principal at IQ Capital): Former diplomat and deeptech VC with a focus on decarbonisation,
Adam Mandel (Entrepreneur in Residence, Carbon 13, and Energised AI co-founder): With over 15 years in applied AI and energy tech,
Prof. Anil Madavapeddi (Planetary Computing, University of Cambridge): Expert in AI for climate and Earth observation.
Bullet Points: Key Takeaways
AI’s Growing Footprint: Training frontier AI models is massively energy-intensive—think billions of pounds and the equivalent annual power consumption of entire UK counties.
Global Competition & UK’s Role: The US and China dominate in sheer model scale and capital, but the UK can lead with small, efficient models, edge computing, and unique research talent.
Climate Shockwaves: A 2.5ºC world would “fundamentally restructure the world’s global supply chain,” warns Prof. Madavapeddi.
Sovereign AI & Infrastructure: Expect a world where every country wants their own AI infrastructure—raising urgent questions about energy independence, data localisation, and technological sovereignty.
Hardware & Efficiency Race: Innovations like IPUs and federated learning promise greater sustainability, but as costs drop, usage surges, so efficiency gains may be offset by soaring demand (Jevons’ Paradox in action).
Quotes to Inspire & Challenge
“If 100 people in this room ran 100 AI tasks daily, their energy use each year would equal the whole of Oxfordshire’s electricity consumption.” — Jono Evans
“At 2.5ºC of warming, the basic mechanisms of the way we live will change fundamentally—especially food security and global supply chains.” — Prof. Anil Madavapeddi
“AI is a tool. Whether it’s good or bad will depend on the human culture that uses it.” — Adam Mandel
Real Talk: AI, Talent, and Global Equity
The conversation dives into tech migration, talent wars, and how AI and climate will influence where the world’s smartest prefer to live and work. While AI enables remote, distributed innovation—even “vibe coding on an anthill in the Okavango Delta”—the race for compute and data sovereignty is heating up.
The panel doesn’t shy from tough questions, including:
Will AI widen the gap for the developing world, especially where diesel and renewables collide for basic connectivity?
Can new federated and edge approaches democratise access, or will ‘AI nationalism’ lead to new inequalities?
Are we at risk of losing decades of institutional knowledge in engineering, or can AI be our bridge to innovation?