Former Nvidia executive says physical AI is still in its building-block era

A broader definition of physical AI

Amulya Vishwanath, founder and managing partner of Techable VC, argues that physical AI should not be treated as a synonym for robotics. In a TechCrunch Equity podcast published October 16, 2026, she described the category as including molecules, materials, medicines and machines on factory floors, alongside robots.

Vishwanath previously spent seven years leading product marketing for Nvidia’s robotics platforms. Her current view is explicitly an investor’s assessment: Techable VC makes early-stage investments across robotics, materials science and life sciences. The TechCrunch item does not present an independent market study or product test.

Why she rejects a single “ChatGPT moment”

Vishwanath characterizes robotics as being in a “Lego era.” Her point is that core components—including actuators, robotic hands, software and safety standards—are still being rebuilt and combined. That leads her to challenge the claim that one general foundation model will soon make much specialized robotics work redundant.

This position contrasts with arguments from some other founders that a general model could create a breakthrough comparable to ChatGPT’s late-2022 launch. TechCrunch frames the disagreement as a debate about where technical progress must occur: in one broadly capable model, or across hardware, software and safety components that must work together in physical settings.

Investment thesis and stated use cases

The podcast also covers what Vishwanath calls “conscious automation.” TechCrunch summarizes her preferred opportunities as responses to concrete constraints, including family-run factories unable to find workers and surgeons experiencing burnout. These examples illustrate her investment thesis; the article does not provide adoption figures or evidence that the proposed systems have solved those problems.

She also discusses the United States’ and China’s respective strengths, her Nvidia-era network as a resource for due diligence and customer introductions, and the need for humanoid robots to demonstrate uses beyond staged performances. The episode offers a practitioner-investor perspective on an emerging field, with conclusions that remain claims and priorities rather than independently verified forecasts.