The end of Moore's Law is the primary catalyst forcing the computing industry to adopt accelerated, specialized architectures over traditional general-purpose CPUs.
The cost of AI is declining at an exponential rate (over 10x annually) and will continue to accelerate, unlocking currently unimaginable applications.
The next major wave of AI value creation will be in vertical-specific applications, such as digital biology and financial trading, and in embodied AI like robotics.
NVIDIA is the central player in the AI revolution, providing a full-stack platform that leads in diverse areas from autonomous vehicle safety to scientific research models.
The demand for AI compute is so significant that it is driving massive, tangible infrastructure projects, including new semiconductor fabs and investment in sustainable energy.
▶The Economics of Exponential AIFeb 2026
Hanson posits that the AI industry is defined by hyper-deflationary economics. He claims the cost of AI is decreasing by over 10x annually due to combined innovations in models, algorithms, and hardware, and predicts this will accelerate to a 100,000x-1,000,000x cost-performance improvement within a decade.
This perspective suggests that the primary investment barrier is not the current cost of computation but the pace of innovation, implying that companies who can best leverage this deflationary curve will dominate future markets.
▶The End of General-Purpose ComputingFeb 2026
According to Hanson, the demise of Moore's Law has triggered an irreversible shift away from general-purpose CPUs toward accelerated computing. This new paradigm is characterized by vertically integrated systems where entire server racks function as a single GPU, as seen in NVIDIA's Grace Blackwell architecture.
Analysts should view the computing industry as bifurcating, with legacy systems losing relevance and value accruing to companies that control the new, specialized, full-stack computing platforms.
▶AI's Next Frontier: Vertical Applications and EmbodimentFeb 2026
Hanson predicts the focus of AI innovation is moving from foundational models to specialized, vertical applications. He foresees major breakthroughs in fields like digital biology, which he believes is on the cusp of a 'ChatGPT moment,' and in robotics, with generative AI enabling advancements in humanoid and multi-embodiment systems.
Investors should look beyond the current large language model race and identify emerging opportunities in industry-specific AI applications and the physical embodiment of AI, which Hanson sees as the next major growth vectors.
▶The Physical Infrastructure of IntelligenceFeb 2026
Hanson emphasizes that the AI revolution is a physical-world phenomenon, not just a software trend. He points to the massive capital investment in new chip fabrication plants by TSMC and SK Hynix and highlights the demand for AI infrastructure as a key driver for the sustainable energy industry.
The AI supply chain, from semiconductor manufacturing to energy production, represents a critical and potentially constrained area for growth, making these sectors a proxy for investing in the broader AI trend.