“An archival 2023 interview with Anthropic co-founder Jared Kaplan discussing scaling laws, large language models, and the trajectory of artificial intelligence.”

“What is the universe? What is intelligence? What is the mind?”

When theoretical physicist turned AI researcher Jared Kaplan shifted his focus to artificial intelligence, he wasn’t just looking to build software—he was trying to answer fundamental questions about reality, perception, and thought itself.

In this episode, we’re revisiting a 2023 conversation with Jared, co-founder of Anthropic, recorded just as he was joining the company.

At the time, we sat down to discuss the physics of intelligence, scaling laws, large language models (LLMs), and where machine intelligence was heading. Looking back three years later, with tools like Claude and ChatGPT transforming the global landscape, listening to this snapshot offers an incredible look at how fast the field has evolved—and how much of today’s reality was already taking shape in those early research models.

We begin with Jared sharing the story of what first drew him from physics into the world of AI, driven by the ultimate quest to understand how mind and universe intersect.

In This Episode, We Cover:

• From Theoretical Physics to AI: How foundational questions about the universe and human cognition led Jared from fundamental physics to machine intelligence.
• The Physics of Scaling Laws: Understanding the empirical patterns that predict how language models improve as computing power, data, and parameters expand.
• Large Language Models & Emergent Abilities: Early insights into how LLMs acquire complex reasoning behaviors.
• The Trajectory of AI Safety & Alignment: Why Anthropic was formed and the core challenges of building safe, cooperative frontier models.

https:/natureandsciencepodcast.com