Cerebras raised $1.1B in a Series G round led by Fidelity to scale manufacturing and data centers amid surging AI demand. Feldman critiques NVIDIA's innovation slowdown, citing strategic financial moves over technical breakthroughs. He challenges assumptions about rapid chip depreciation, highlighting Cerebras’ wafer-scale architecture that overcomes memory bottlenecks. The inference market is expanding faster than training due to lower software barriers. Energy demands for AI raise sustainability concerns, with nuclear power emerging as a viable solution. Market concentration in the 'Mag7' stocks poses hidden risks, while talent shortages—worsened by U.S. immigration policies—are a critical bottleneck. Despite high valuations, supply constraints persist, with demand far outpacing capacity. The U.S. must improve education, immigration, and energy policy to compete with China. Much of Cerebras’ revenue comes from the UAE, reflecting global shifts in AI infrastructure driven by sovereignty and investment. True innovation, Feldman argues, requires relentless dedication, not work-life balance.