AI This Week: NVIDIA’s Record Revenue, Elon’s Data Centers in Space & Gemini 3’s Insane Performance w/ Salim Ismail, Dave Blundin & Alexander Wissner-Gross | EP #210
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
This Moonshots podcast episode explores how we're witnessing the "speed running" of science fiction concepts into reality. (00:00) The hosts discuss NVIDIA's record-breaking revenues and position as the "central bank for AI," Elon Musk's plans for space-based AI infrastructure generating 100 gigawatts annually, and the convergence of multiple exponential technologies from robotics to energy production. (00:40) They examine how major tech companies are forming unprecedented partnerships, building massive data centers, and preparing for a future where AI becomes the largest consumer of electricity in human history.
- Core themes include NVIDIA's dominance and future challenges, space-based computing infrastructure, energy abundance through multiple technological pathways, and the acceleration of robotics deployment across industries
Speakers
Peter Diamandis
Peter Diamandis is a renowned entrepreneur, author, and founder of the XPRIZE Foundation. He serves on the board at FII (Future Investment Initiative) and regularly advises on emerging technologies and exponential trends. He's the co-author of multiple bestselling books including "The Future Is Faster Than You Think" and the upcoming "We Are As Gods: Survival Guide for the Age of Abundance."
Dave Blundin
Dave Blundin is the founder and General Partner of Link Exponential Ventures, a Boston-based venture capital firm. He focuses on early-stage investments in exponential technologies and has extensive experience in identifying breakthrough innovations in AI, robotics, and other emerging sectors.
Salim Ismail
Salim Ismail is the founder of OpenExO and a leading expert on exponential organizations. He helps companies and governments understand and implement exponential business models and technological transformation strategies.
Dr. Alexander Wissner-Gross
Dr. Alexander Wissner-Gross is a computer scientist, physicist, and founder of Reified. He holds advanced degrees in physics and computer science and is known for his work on artificial intelligence, computational sustainability, and complex systems research.
Key Takeaways
NVIDIA Has Become the Central Bank of AI
NVIDIA's transformation from a graphics card company to the backbone of AI infrastructure represents a fundamental shift in how we think about technological monopolies. (00:48) The company reported $57 billion in revenue with 62% year-over-year growth, effectively "minting their own currency" in the form of compute power. (06:54) This positions them uniquely as every AI company must "buy their currency" to access the computational resources needed for training and inference. However, this dominance faces challenges from Google's TPUs, AMD, and specialized AI chips designed for specific algorithms.
Space-Based Computing Will Transform Energy Bottlenecks
Elon Musk's vision of deploying 100 gigawatts per year of AI infrastructure in orbit represents a paradigm shift from Earth's power grid constraints to launch capacity constraints. (27:31) The first H-100 chip successfully deployed to space demonstrated that radiative cooling using aluminum can work effectively, opening the door to massive space-based data centers. (37:00) This approach leverages free and abundant solar energy in space while distributing compute globally at the speed of light, potentially making orbital computing cheaper per unit than terrestrial alternatives.
AI Will Solve Biology and Medicine Within This Decade
Anthropic's aggressive hiring of life science researchers signals the convergence of AI and biological sciences reaching a critical inflection point. (81:58) Dario Amodei has publicly stated his expectation that disease, biology, and medicine will be "solved" by the end of the decade. This prediction aligns with AI systems like Gemini 3 already outperforming radiological trainees and approaching board-certified radiologist performance levels. (82:43) The combination of AI diagnostics and epigenetic reprogramming entering human trials in 2026 suggests we're on the cusp of treating aging itself as a reversible condition.
Robotics Training Through Human Demonstration Is Unlocking Mass Deployment
The breakthrough in robotics comes from training methods that eliminate traditional programming barriers. (71:04) Companies like Sunday Robotics have created armies of "memory developers" who wear specialized gloves that record every human action, building the world's largest dexterity datasets. This approach, combined with vision-language-action models, makes robot training as simple as showing them what to do rather than coding complex instructions. (73:02) This democratization of robot training will accelerate deployment across industries, from dangerous nuclear reactor work to space-based construction.
Energy Abundance Is Becoming Reality Through Multiple Pathways
The convergence of solar, nuclear, and space-based energy production is creating unprecedented energy abundance. (42:03) China's massive lead in electricity generation demonstrates what's possible when countries commit to energy transformation, while technologies like pebble bed nuclear reactors offer melt-down proof alternatives to traditional nuclear power. (54:54) The key insight is that AI is driving demand for energy so aggressively that it's pulling forward all forms of energy production, creating a positive feedback loop where energy abundance enables more AI development, which in turn drives more energy innovation.
Statistics & Facts
- NVIDIA reported $57 billion in revenue with 62% year-over-year growth and expects $65 billion in revenue next quarter. (05:30) This represents their transformation from a graphics card company to the dominant force in AI infrastructure.
- Elon Musk plans to deploy 100 gigawatts per year of AI infrastructure in orbit, equivalent to roughly a quarter of total US electricity output (which averages 460 gigawatts). (28:42) This would be equivalent to adding 100 nuclear power plants worth of capacity annually.
- China is producing approximately 325 terawatts of electricity versus 80 terawatts for the US, showing a roughly 4x advantage in total energy generation. (42:42) China leads in solar, wind, nuclear, and gas generation simultaneously.
