Yann LeCun has long argued that current LLM-centric approaches are a dead end for AGI, advocating instead for world models grounded in richer representations of reality. With his recent decision to build his next venture in Paris rather than Silicon Valley—alongside the rise of companies like Mistral and Hugging Face—Europe seems to be gaining credibility as a place for foundational AI research. Is this an isolated personal move, or an early signal of a broader Silicon Valley → Europe shift driven by research culture, regulation, and geopolitics? What would need to change for Europe to sustain such a transition?