AI economy hits physical limits: chip shortages, energy constraints, and a new paradigm

During the Milken Global Conference held in Beverly Hills, five leaders from the AI supply chain addressed existing bottlenecks. Christophe Fouquet, the CEO of ASML, cautioned that chip production will face supply constraints for the next two to five years, impacting major hyperscalers such as Google, Microsoft, Amazon, and Meta. Francis deSouza, COO of Google Cloud, shared that the division's revenue exceeded $20 billion last quarter, marking a 63% increase, with a backlog totaling $460 billion, and mentioned the exploration of orbital data centers. Qasar Younis, CEO of Applied Intuition, highlighted the importance of real-world data for physical AI, linking it to national sovereignty. Dimitry Shevelenko from Perplexity discussed their transformation into a "digital worker" platform, while Eve Bodnia of Logical Intelligence is working on faster energy-based models. The panel conveyed a positive outlook on AI's capabilities across various sectors.