In detail
- Extreme concentration: top ten US stocks now represent 40% of S&P 500 market cap (vs. 17% in 2015).
- Nvidia's AI accelerator share declining from 85% (2023) to estimated 75% (2026); custom chips from Google and Amazon cut operating costs by 30–40%.
- OpenAI and Anthropic face profitability risks: fast revenue growth but massive compute costs; rising token prices push companies toward cheaper open-source models.
- Leveraged chip ETFs have quintupled their influence on global markets since early 2024.
Why it matters
The warning signals market distortions making AI infrastructure and service investments risky. For German SMEs, this means: caution on expensive frontier models—cheaper open-source alternatives will reach parity faster.
For you Audit whether your AI strategy relies on costly frontier models; cheaper open-source alternatives may soon match performance at a fraction of the cost.