Lior Susan says physical AI will replace husbands and dominate public markets

For a decade, the venture capital world obsessed over the infinite scalability of software, leaving the gritty reality of manufacturing and hardware in the rearview mirror. Lior Susan, co-founder of Eclipse Ventures, argues that this era of 'vibe coding'—building digital solutions with no physical moat—is hitting a wall. While software margins are attractive, 85% of global GDP remains trapped in physical industries like mining, defense, and manufacturing. The tide is turning. Investors are realizing that while you can easily replicate a SAS platform, you cannot easily replicate a semiconductor clean room or a global satellite network.

Physical AI and the End of Domestic Labor

The most provocative claim in the current tech cycle isn't about chatbots; it’s about the imminent arrival of functional household robotics. Susan predicts that by the end of 2026, consumers will be able to purchase a home robot for roughly $5,000—the price of a high-end washing machine—capable of autonomous task execution. These machines won't require manual mapping. Instead, they use Physical AI and reinforcement learning to navigate homes, sort laundry, and clean kitchens. Susan jokingly suggests these machines might even render husbands obsolete, highlighting a massive shift where robots transition from rigid industrial cages to dynamic human environments.

Geopolitics and the Five Forces of American Building

The resurgence of hardware isn't just a technological trend; it's a geopolitical necessity. Lior Susan identifies a convergence of five forces—capital, talent, policy, customer demand, and technology—that haven't aligned in the U.S. since the era of Henry Ford. As deglobalization forces manufacturing back to domestic shores, the demand for automation and energy storage has skyrocketed. This shift is reflected in the massive capital influx into the Eclipse Ventures portfolio, which raised $4.5 billion in equity in just the first quarter of 2026, surpassing the total raised during the firm's first eight years of existence.

The SpaceX IPO and the Multiples Game

SpaceX stands as the ultimate testament to the profitability of 'atoms.' As the company nears a historic IPO with valuations potentially crossing $2 trillion, Elon Musk is masterfully blending hardware dominance with AI speculation. Susan notes that Musk is likely leveraging the massive cash flows from Starlink to finance xAI. By integrating Physical AI into a hardware-heavy asset, Musk secures the high valuation multiples typically reserved for software while maintaining the defensible moat of physical infrastructure. This strategy forces the public market to value real assets and EBITDA over the vanity metrics of gross margins.

Efficiency Gains via Transformer Models

While hardware has traditionally been capital-intensive, the advent of transformer models is radically lowering the cost of entry. Companies like Wayve are achieving autonomous driving milestones previously pioneered by Waymo but with a fraction of the capital. By using synthetic data and transfer learning, these startups bypass the need for massive, expensive physical testing fleets. This 'capital-light' approach to 'heavy' industries allows new challengers to disrupt incumbents, proving that while atoms are the goal, bits are the fuel that accelerates the journey.