AI Hype Cycle Continues
AI hype is still dominating news cycles: Yann LeCun’s AMI raised $1B+, NVIDIA eyes gigawatt-scale compute, and GPT-5.4 sets benchmarks [https://x.com/i/status/2031588163823468804] [https://x.com/i/status/2031963245527838883].
Yann LeCun's new venture, Advanced Machine Intelligence (AMI) Labs, secured $1.03 billion in seed funding to develop AI systems using "world models". This approach aims to enable AI to reason, plan, and understand real-world environments, differing from the current reliance on large language models. The funding round values AMI at $3.5 billion. NVIDIA is planning a major shift to 800-volt DC power architecture for data centers to support the increasing power demands of AI. They're investing $2 billion into Nebius to deploy 5 gigawatts of NVIDIA systems by 2030, creating AI factories and hyperscale clouds. This move addresses the limitations of low-voltage systems and prepares for future AI infrastructure needs. GPT-5.4 has set new benchmarks in AI performance, particularly in computer use and professional applications. It achieved a 75% accuracy on the OSWorld benchmark, surpassing human performance in desktop computer operation. On the GDPval benchmark, GPT-5.4 scored 83%, matching or exceeding industry experts across 44 professions.
