Search for a command to run...

Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
In this episode of Moonshots, Peter Diamandis joins Dave Blundin and Dr. Alexander Wissner-Gross for an in-depth analysis of the latest tech developments. Peter opens by reporting from XPRIZE Visioneering 2025, where three groundbreaking challenges received funding: the Abundance XPRIZE (delivering essential services for $250/month), a Fusion XPRIZE, and the WALL-E prize for automated waste sorting. (12:00)
Founder and Executive Chairman of XPRIZE, Chairman of Singularity University, and founder of over 20 companies. He chairs the AI activities at Saudi Arabia's Future Investment Initiative and recently led XPRIZE Visioneering 2025, securing $3.5 million in funding for breakthrough challenges.
Founder and General Partner of Link Ventures, with extensive experience founding over 20 direct-to-consumer AI companies. He serves on the XPRIZE board and specializes in identifying AI market opportunities and competitive positioning strategies.
Computer scientist, physicist, and founder of Reified, focused on AI and complex systems. Former researcher at MIT and Harvard, he provides deep technical analysis on superintelligence, quantum computing, and the convergence of AI with physical world applications.
OpenAI's launch of the Atlas browser isn't just another product - it's a strategic move to control the distribution channel for superintelligence. (13:45) As Dave explains, this mirrors Bill Gates' historical strategy of controlling user entry points. While the browser functionality may seem basic now, it positions OpenAI to capture user data and create switching costs. The integrated ChatGPT experience, browser memory, and agent capabilities represent OpenAI's attempt to build sustainable competitive advantages beyond just having the best AI model.
The convergence of AI with biology and chemistry is creating "science data factories" that operate 24/7 to mine new knowledge from nature. (26:16) Companies like LILA are using superintelligence models to generate scientific hypotheses, then deploying robotic systems to test them continuously. This represents a fundamental shift from human-limited research cycles to AI-driven discovery that can potentially double human lifespan within the decade, as suggested by Anthropic's leadership.
The path to superintelligence is fundamentally constrained by energy production capacity. (46:00) Eric Schmidt's projection that we need 100 gigawatts by 2030 - just for AI compute - represents only a 10% expansion of US power supply, but the exponential growth continues. The timeline convergence of fusion power (targeting 2028-2030), SMR deployment, and peak AI compute demand will determine which nations and companies dominate the superintelligence era.
While Google's Willow chip demonstrates quantum advantage in specific problems, the field still lacks economically transformative applications. (74:00) The real breakthrough will come when quantum systems can accelerate AI training or inference by orders of magnitude. Currently, quantum remains relegated to niche applications while classical AI systems continue solving problems that were once thought to require quantum computation, like protein folding.
From Amazon's delivery glasses to Tesla's unified A15 chip architecture, AI is finally walking out of data centers into the physical world. (68:00) These aren't just productivity tools - they're data collection systems training the next generation of autonomous agents. Every human wearing AI-enabled devices becomes a trainer for robotic systems, creating massive feedback loops that will accelerate automation across the entire service economy.