Toronto
Hosted by Trajectory Labs
Explore how critical minerals shape AI infrastructure, geopolitical power and Canada’s future at this Trajectory Labs talk in Toronto. On Tuesday 15 September 2026, from 6:15pm to 9:00pm EDT, Trajectory Labs presents a detailed journey through the physical supply chains behind frontier AI. While discussion of AI often focuses on models and software, this event examines the gallium, germanium, rare earths, copper, tungsten and other materials required to manufacture GPUs, servers, cooling systems and data-centre power infrastructure. From mine to data centre Noah Chaikin will map the AI value chain from mineral extraction and refining through to compute infrastructure. The talk will examine how production concentrated in a small number of countries creates strategic vulnerabilities, and how export restrictions on gallium, germanium, graphite and rare earth magnets demonstrate the use of supply-chain chokepoints as instruments of statecraft. Key questions include: - What does AI infrastructure physically consist of? - Which critical minerals does the AI industry depend on? - Where is control of mining and refining capacity heading? - How could North America build more secure mineral supply chains? - What might the convergence of mining, refining and AI investment mean for Canada over the next decade? Chaikin brings a foresight perspective informed by his work on geopolitical statecraft at Kearney Foresight and his research into Canadian rare earth supply chains. He will also consider Pax Silica, the US-led initiative to develop a trusted AI supply chain among allies, as a live example of governments responding to these dependencies. The session closes with scenarios for how the sector could evolve and the implications for Canada. Who should attend? This talk is relevant to people working in or studying AI infrastructure, technology policy, governance, geopolitics, mining, energy, industrial strategy and supply-chain resilience. It should also appeal to builders and researchers who want to understand the material systems beneath advanced AI, not only its software layer. No specific technical prerequisites are listed.… Register for this event





