Y Combinator

We're entering a new era of software where a single person, working with AI agents, can build products that previously required entire teams. In this episode of Lightcone, the hosts break down the rise of AI coding agents, "tokenmaxxing", and the emerging workflows behind tools like Claude Code and OpenClaw. They discuss why AI systems today feel less like productivity tools and more like collaborators, why the future of AI should be personal and user-controlled, and how founders are starting to build software in completely new ways.

Paul Graham, the Founder of Y Combinator, speaks live from Stockholm Waterfront Congress Centre in Stockholm, Sweden. He covers two topics -- if you should you go to Silicon Valley and what Sweden can do to thrive as a Startup hub.

We're entering a new era of software where a single person, working with AI agents, can build products that previously required entire teams. In this episode of Lightcone, the hosts break down the rise of AI coding agents, "token maxxing", and the emerging workflows behind tools like Claude Code and OpenClaw. They discuss why AI systems today feel less like productivity tools and more like collaborators, why the future of AI should be personal and user-controlled, and how founders are starting to build software in completely new ways.

Razorpay founder Harshil Mathur shares how a side project turned into one of India’s biggest payment platforms. He talks about starting as a coder with no fintech background, discovering that online payments in India were painfully hard, and slowly finding the right customer by talking directly to startups. The biggest lessons are about trust, speed, and staying close to the problem. In B2B, he says, customers do not just buy features — they buy reliability and human support. He also explains why regulated markets are hard, but that difficulty can become a serious moat.

Demis Hassabis argues current AI methods are part of the AGI path, but continual learning, memory, long-term reasoning, and consistency remain unsolved. He expects AGI around 2030, making agents, distilled models, multimodal Gemini, and open Gemma important. He highlights AlphaFold’s scientific impact, future virtual cell modeling, and urges founders to pursue interdisciplinary deep tech with clear objectives, strong data or simulators. True creativity and novel scientific hypothesis generation remain open challenges.

In this episode, Kuan Vuong of Physical Intelligence explores why robotics may be reaching a true inflection point. He explains how recent advances in foundation models, cross-embodiment learning, and real-world deployment are making robots more adaptable, more useful, and faster to build than before. Using examples like laundry folding and warehouse packaging, the conversation shows how robotics is shifting from a difficult, highly customized engineering problem into a more scalable platform opportunity. Overall, it is a clear and forward-looking discussion about why the cost of building in robotics is falling, what mixed autonomy looks like in practice, and how this could unlock a new wave of robotics startups.

This episode explores how Billion2One grew from a bold scientific idea into one of the fastest-scaling molecular diagnostics companies in the U.S. Founders Oguzhan Atay and David Tsao share how they built a breakthrough blood-based testing platform that can detect extremely rare DNA signals with remarkable precision, first in prenatal screening and now in cancer diagnostics. The conversation traces their journey from two PhD students with limited resources to a public company processing hundreds of thousands of tests each year, while also looking ahead to their bigger ambition: building ultrasensitive blood tests that could one day detect cancer at its earliest, most treatable stage.