Five architects of the AI economy explain where the wheels are coming off
AI leaders discuss real-world limitations in chip supply, energy, data, and fundamental architecture.Summary
Five key figures in the AI economy, including leaders from ASML, Google Cloud, Applied Intuition, Perplexity, and Logical Intelligence, convened at the Milken Global Conference to discuss the significant challenges and limitations facing the AI boom. They highlighted real physical constraints such as chip shortages, which are expected to persist for years, and the immense energy demands of AI, prompting Google Cloud to explore orbital data centers. The discussion also touched upon the bottleneck of real-world data collection for training AI systems, the potential of alternative AI architectures like energy-based models (EBMs) that mimic human reasoning, and the evolving role of AI agents in tasks and security. Geopolitical implications, particularly the entanglement of physical AI with national sovereignty and the limitations faced by countries like China due to access to advanced chip manufacturing, were also addressed. Finally, the panelists offered optimistic perspectives on how AI could empower future generations by enabling them to tackle complex problems and fill labor shortages, rather than diminishing critical thinking.
(Source：TechCrunch)