Interview with Brett Adcock, Serial Entrepreneur Worth $19 Billion: Commercialization Turning Point for Figure, New Lab HARK Creating a Revolution in AI Hardware and Computer Use

Podcast interview titled "$39B founder: “AI will eat the whole internet”" (My First Million podcast featuring Brett Adcock, founder of humanoid robot unicorn Figure, and former founder of Archer Aviation / Vettery). Here are the key points summarized: 1. Journey as a serial entrepreneur and the "Binary" business philosophy • From an Illinois farm to a $10 billion empire: • Founded and sold Vettery for over $100 million, took electric aircraft company Archer Aviation public; then invested personal assets, even mortgaged property to establish humanoid robot company Figure (valued at hundreds of billions), and incubated the campus gun prevention project Cover and the new generation AI lab HARK. • Binary Outcome: • No obsession with personal paper wealth (estimated by Fortune at about $19 billion); driven by extreme "conquest desire and ambition to reach the summit." • Firm belief that the outcome of robotics and hardcore technology is extremely binary—either it scales to change human productivity towards a trillion-dollar market value or it goes to zero, with no mediocre middle ground. 2. New lab HARK: Breaking API limitations to achieve true "AI Computer Use" • 99.9% of internet scenarios lack APIs: • Less than 1 in 1000 of the world's 1000 websites provide comprehensive Consumer APIs. Relying solely on APIs or MCP cannot build a true universal digital assistant (like ordering from DoorDash, complex ticketing, internal system operations). • HARK's technical path: Independently launching systems through cloud-based virtual sandboxes, using self-developed post-training reinforcement learning algorithms, allowing AI to control the entire internet like humans through "visual recognition of screens + moving the mouse + typing on the keyboard." • New AI-native hardware that "kills phones and computers": • Existing touch-screen-based iPhones and MacBooks were designed 20 years ago, not for the AI era. • HARK has assembled a team led by former core hardware design leaders from Apple, secretly developing the next generation of AI-native hardware, completely breaking free from the App Store framework, building all-weather multimodal interaction, edge-specific weights, and lifelong memory digital Jarvis. 3. Commercialization of Figure humanoid robots and general household bottlenecks • B2B industrial logistics scenarios are the first to close the loop (BMW case): • The automotive and logistics manufacturing sectors face over 100% turnover rates and labor exhaustion. Figure robots have completed real-world tests sorting single packages in 2.9 seconds and running continuously for over 200 hours, surpassing human workers in speed and ROI. • The real challenge of transitioning from industrial to household (General Robotics): • The difficulty lies not in manufacturing hardware but in the generalization of embodied intelligent brains in unstructured new environments. • For example, recognizing and folding clothes of different lighting, heights, and materials in homes never visited before. Figure is currently focused on tackling end-to-end large model generalization from "pixels to torques." 4. Hardcore entrepreneurial rules: Doing extremely difficult things is actually easier • Non-linear return logic: • Building a quadruped robotic dog may be one-third the difficulty of humanoid robots, but the commercial TAM for robotic dogs is very small; while humanoid robots, despite being 3-4 times harder, have a return rate that is in the millions. • Tackling the hardest hardcore problems naturally filters out 99% of mediocre competitors while attracting the world's top surplus talents and capital. • Technical interviews and talent identification: • 90% of engineers in the Bay Area cannot handle true hardcore breakthroughs. Personally conducting in-depth technical evaluations of core personnel to discern candidates as "practitioners who have experienced deep pits and bear scars" versus "theorists who have only observed others' achievements." 5. Personal workflow, energy allocation, and guidance from top mentors • The "three-drawer" rule of life: • Dividing life into "career, family, and social matters." To achieve A+ in career and family (with three children), he completely cut out all meaningless socializing, golf, and long-distance parties. • Inspiration from Jeff Bezos and Jensen Huang: • Bezos, as an investor in Figure, advised him: "You have completed the hardest part from 0 to 1; the next 1-2 years will be a life-and-death battle to truly scale the robots to mass production in the industrial sector, this is a decisive battle." • Firm belief that within the next 10-20 years, AI will consume the entire physical and digital world at a scale 100 times larger than the internet.