The method of invention, AI's new clock speed and why capital markets are confused - Azeem Azhar's Exponential View Recap
Podcast: Azeem Azhar's Exponential View
Published: 2025-12-05
Duration: 24 min
Summary
The episode examines the rapid evolution of AI technologies, particularly focusing on the advancements by OpenAI and Google, and the challenges faced by capital markets in adapting to these exponential changes.
What Happened
ChatGPT marks its third anniversary with nearly 900 million users, showcasing its widespread adoption and daily usage by a third of its user base. This highlights the massive scale and demand for compute power that large language models require, which seems to overshadow other technological advancements.
OpenAI's attempt to supersede GPT-4 with GPT-4.5 fell flat, but a pivotal shift occurred with the introduction of reasoning models like 01 and 03 that excel in inference tasks. Google's Gemini Pro exemplifies a blend of large language and reasoning models, hinting at a new underlying technology.
NVIDIA's GPUs have outpaced Moore's Law, doubling AI throughput over the past decade, demonstrating the accelerated clock speed of AI development. These advancements demand cutting-edge technologies to manufacture complex chips, emphasizing the rapid technological evolution.
AI is a general-purpose technology that reduces costs and creates new behaviors, such as leveraging AI for contract reviews and personalized task management. These tools have significantly improved productivity, allowing users to build reliable software tailored to their needs.
The episode delves into the complexities of capital markets struggling to fund exponential technologies like AI. Traditional market valuations fail to capture the expansive and diffuse effects of these technologies across industries.
Investors are challenged by the unpredictability of profit allocations with AI. While new technologies expand markets, it remains unclear where profits will concentrate, drawing parallels to historical technology shifts like the internet and internal combustion engines.
The software development market is poised for transformation, with AI enabling individuals to create software independently. This raises questions about the future size and nature of this market, as well as the impact on white-collar jobs.
Capital markets' current frameworks are ill-suited for the exponential growth and system-wide improvements brought by AI, highlighting the need for a paradigm shift in how financial markets approach technology investments.
Key Insights
- ChatGPT has reached nearly 900 million users in three years, with a third of its user base engaging daily, indicating the substantial demand for the compute power required by large language models.
- NVIDIA's GPUs have surpassed Moore's Law by doubling AI throughput over the past decade, necessitating advanced technologies for manufacturing increasingly complex chips.
- AI's role as a general-purpose technology is evident in its ability to reduce costs and foster new behaviors, such as automated contract reviews and personalized task management, significantly enhancing productivity.
- Traditional capital markets struggle to evaluate exponential technologies like AI, as they fail to account for the widespread and diffuse impacts across various industries, necessitating a new approach to technology investments.