We are living in an ExPro world — exponential and probable — where prediction is futile and paradoxes abound. The rise of Agentic AI has transformed the debate between Frontier Models like Anthropic’s Mythos and OpenAI’s GPT‑5.6, and the fast‑advancing Open Source models from China such as DeepSeek and Kimi. Scaling Laws continue to hold, driving breakthroughs in inference and training, while the global AI buildout has triggered trillion‑dollar investments in data centers and Nvidia GPUs. Yet, black markets thrive, with chips like the H100 and B200 commanding multiples of their official prices in China.
As enterprises chase premium AI tokens and governments wrestle with bans and distillation campaigns, the question remains: can Frontier AI companies sustain profitability when Open Source alternatives catch up within months? From Rothschild’s Waterloo advantage to Jim Simons’ Medallion Fund, history shows that intelligence — not just information — creates fortunes. In this brave new world, the only certainty is uncertainty.