Nvidia Veteran Amulya Vishwanath on Why Robotics is Still in the 'Lego Era'

The 'Lego Era' of Robotics

Amulya Vishwanath, a seasoned veteran who spent seven years leading product marketing for NVIDIA’s robotics platforms, offers a grounded perspective on the current state of the industry. While mainstream excitement focuses on the imminent arrival of general-purpose foundation models for machines, Vishwanath argues that robotics is currently in what she describes as its "Lego era."

This phase is characterized by a fundamental rebuilding of the sector's basic building blocks. According to Vishwanath, essential components such as actuators, robotic hands, specialized software stacks, and safety standards are still being refined and restructured. This suggests that the industry is not yet at a plug-and-play stage where software can simply be downloaded into a universal frame to perform complex tasks. Instead, hardware and software development must still move in lockstep to solve mechanical and safety constraints.

Redefining Physical AI

One of the key takeaways from Vishwanath’s transition from Nvidia to the investment side is a broader definition of "Physical AI." While humanoid robots often dominate the headlines and capture the public imagination, the scope of physical intelligence is significantly wider. Physical AI refers to any system where artificial intelligence interacts directly with the physical world to perform tasks or interpret environments.

This includes industrial automation, autonomous logistics, and specialized machinery that may not resemble a human or even a traditional robot. For founders, this distinction is critical: the value proposition often lies in solving specific physical bottlenecks rather than pursuing general-purpose robotics. The challenge for startups today is navigating this "Lego era" by building reliable hardware that can actually leverage the emerging intelligence of modern AI models.

Implications for Investors and Founders

For investors, Vishwanath’s perspective serves as a reminder that the robotics stack is still being defined. The "building blocks"—from the way a robot senses its environment to the way it moves—represent significant investment opportunities. We are seeing a shift from pure software plays to integrated systems that require deep expertise in materials science and mechanical engineering.

Founders entering this space must be prepared for a longer development cycle than pure-play AI software startups. Success in the current landscape requires a focus on the foundational layers of the robotic stack, ensuring that the physical components are as robust and sophisticated as the neural networks driving them. As the industry matures beyond the Lego era, these early architectural choices will likely determine which platforms become the standards of the future.