Why Physical AI is Entering Its ‘LEGO Era’ According to Amulya Vishwanath

The robotics industry is standing on the precipice of a transformation that mirrors the early days of generative software. According to Amulya Vishwanath, a former Nvidia strategist turned venture capitalist, Physical AI is currently moving through its ‘LEGO era.’ Just as the building blocks of early digital intelligence once felt modular and experimental, the hardware and software layers of modern robotics are finally beginning to click together in ways that promise widespread utility.

The ChatGPT Moment for Robotics

For years, the promise of autonomous machines has been hindered by siloed development cycles and brittle code. Vishwanath argues that we are rapidly approaching the robotics equivalent of the ‘ChatGPT moment.’ This shift is characterized by foundational models that allow machines to interact with the physical world with a level of fluidity previously reserved for science fiction. The goal is no longer just to program a robot to repeat a task, but to teach it to understand the environment well enough to navigate chaos.

Why the LEGO Analogy Matters

The ‘LEGO era’ refers to the increasing modularity of these systems. Developers no longer need to build an entire operating system from scratch for every single movement. Instead, they are stacking pre-trained models—vision, locomotion, and manipulation—to build complex robotic behaviors. This democratization of capability is lowering the barrier to entry, inviting a new wave of startups to solve real-world logistical and manufacturing challenges that have remained stubborn for decades.

As global supply chains shift and labor shortages persist, the transition from rigid automation to flexible, intelligent agents is becoming a economic imperative rather than a luxury. Vishwanath’s perspective suggests that the winners of this cycle won’t just be the hardware manufacturers, but the platforms that best facilitate this modular, intelligent assembly.

Source: This article was originally reported by TechCrunch and independently adapted and contextualized for our international English-speaking audience. Read the original report here.