Robots are waiting for a ChatGPT moment: Nvidia’s Les Karpas explains why at TechCrunch Disrupt 2026

The core problem in robotics development is the lack of a universal physical world dataset, which directly slows down the creation of versatile robots.

Nvidia's Global Head of Physical AI, Les Karpas, will discuss the significant hurdles preventing robotics from achieving a breakthrough moment akin to ChatGPT at TechCrunch Disrupt 2026. Karpas will explain that unlike language AI, which benefited from vast internet-wide datasets, physical AI for robotics lacks a comparable large-scale, real-world dataset. He will highlight this data gap as the primary bottleneck, even for companies like Waymo in the self-driving car sector, which have accumulated data over years of operation. The session will explore current efforts by startups to artificially generate this data through simulation and synthetic means, and Nvidia's perspective on bridging the digital and physical divide. Karpas, with his extensive background across various industries and roles, is positioned to offer unique insights into the challenges and opportunities within physical AI.

The absence of an internet-scale dataset for physical interactions presents a fundamental challenge for robotics, contrasting sharply with the data-rich environment that propelled large language models. This deficiency means robots cannot learn and generalize behaviors as effectively as their AI counterparts, limiting their potential for widespread adoption and complex task execution. Karpas's upcoming talk at TechCrunch Disrupt 2026 aims to illuminate this critical bottleneck and discuss potential pathways to overcome it.

Startups are actively pursuing solutions, leveraging simulation, synthetic data generation, and foundation models trained across diverse robotic platforms to bridge the digital-physical divide. Nvidia, through its Inception program, is deeply involved in nurturing this ecosystem, suggesting a strategic push to standardize and scale physical AI data collection and utilization. The success of these efforts will determine if and when robots can achieve the transformative impact seen in the language AI sector.