Overview
- The Wall Street Journal reports transformer supplies are severely constrained with capacity booked through 2028, and GE Vernova’s CEO says meeting U.S. power needs for AI will likely take 10 to 15 years.
- Meta executive Alex Schultz defended the company’s AI infrastructure spend of about $72 billion this year, arguing the investment is aggressive yet not out of control and is already improving ads and content recommendations.
- Morgan Stanley finds 15% of S&P 500 companies now cite quantifiable AI-driven gains as of Q3 2025, with technology firms leading the pack.
- AlphaSense data shows 418 U.S.-listed companies worth over $1 billion have added AI-related risk factors to SEC filings this year, up 46% from 2024.
- Elon Musk outlined a plan to launch solar-powered AI satellites targeting roughly 100 gigawatts of capacity as a potential low-cost energy source for large-scale AI, and said Neuralink has implanted devices in more than 10 patients who can now communicate at near normal speed.