ARTIFICIAL INTELLIGENCE

“The U.S. led the software era, but AI is software that you don’t ‘write’-you teach it. Europe can fuse its industrial capability with AI to lead in Physical AI and robotics. This is a once-in-a-generation opportunity.” – Jensen Huang – CEO, Nvidia

In a compelling dialogue at the World Economic Forum Annual Meeting 2026 in Davos, Switzerland, Nvidia CEO Jensen Huang articulated a transformative vision for artificial intelligence, distinguishing it from traditional software paradigms and spotlighting Europe’s unique position to lead in Physical AI and robotics.^1,2,4 Speaking with World Economic Forum interim co-chair Larry Fink of BlackRock, Huang emphasised AI’s evolution into a foundational infrastructure, driving the largest build-out in human history across energy, chips, cloud, models, and applications.^2,3,4 This session, themed around ‘The Spirit of Dialogue,’ addressed AI’s potential to reshape productivity, labour, and global economies while countering fears of job displacement with evidence of massive investments creating opportunities worldwide.^2,3

The Context of the Quote

Huang’s statement emerged amid discussions on AI as a platform shift akin to the internet and mobile cloud, but uniquely capable of processing unstructured data in real time.² He described AI not as code to be written, but as intelligence to be taught, leveraging local language and culture as a ‘fundamental natural resource.’^2,4 Turning to Europe, Huang highlighted its enduring industrial and manufacturing prowess – from skilled trades to advanced production – as a counterbalance to the US’s dominance in the software era.⁴ By integrating AI with physical systems, Europe could pioneer ‘Physical AI,’ where machines learn to interact with the real world through robotics, automation, and embodied intelligence, presenting a rare strategic opening.^4,1

This perspective aligns with Huang’s broader advocacy for nations to develop sovereign AI ecosystems, treating it as critical infrastructure like electricity or roads.⁴ He noted record venture capital inflows – over $100 billion in 2025 alone – into AI-native startups in manufacturing, healthcare, and finance, underscoring the urgency for industrial regions like Europe to invest in this infrastructure to capture economic benefits and avoid being sidelined.^2,4

Jensen Huang: Architect of the AI Revolution

Born in Taiwan in 1963, Jensen Huang co-founded Nvidia in 1993 with a vision to revolutionise graphics processing, initially targeting gaming and visualisation.⁴ Under his leadership, Nvidia pivoted decisively to AI and accelerated computing, with its GPUs becoming indispensable for training large language models and deep learning.^1,2 Today, as president and CEO, Huang oversees a company valued in trillions, powering the AI boom through innovations like the Blackwell architecture and CUDA software ecosystem. His prescient bets – from CUDA’s democratisation of GPU programming to Omniverse for digital twins – have positioned Nvidia at the heart of Physical AI, robotics, and industrial applications.⁴ Huang’s philosophy, blending engineering rigour with geopolitical insight, has made him a sought-after voice at forums like Davos, where he champions inclusive AI growth.^2,3

Leading Theorists in Physical AI and Robotics

The concepts underpinning Huang’s vision trace to pioneering theorists who bridged AI with physical embodiment. Norbert Wiener, father of cybernetics in the 1940s, laid foundational ideas on feedback loops and control systems essential for robotic autonomy, influencing early industrial automation.⁴ Rodney Brooks, co-founder of iRobot and Rethink Robotics, advanced ’embodied AI’ in the 1980s-90s through subsumption architecture, arguing intelligence emerges from sensorimotor interactions rather than abstract reasoning – a direct precursor to Physical AI.⁴

• Yann LeCun (Meta AI chief) and Andrew Ng (Landing AI founder) extended deep learning to vision and robotics; LeCun’s convolutional networks enable machines to ‘see’ and manipulate objects, while Ng’s work on industrial AI democratises teaching via demonstration.⁴
• Pieter Abbeel (Covariant) and Sergey Levine (UC Berkeley) lead in reinforcement learning for robotics, developing algorithms where AI learns dexterous tasks like grasping through trial-and-error, fusing software ‘teaching’ with hardware execution.⁴
• In Europe, Wolfram Burgard (EU AI pioneer) and teams at Bosch/ Siemens advance probabilistic robotics, integrating AI with manufacturing for predictive maintenance and adaptive assembly lines.⁴

Huang synthesises these threads, amplified by Nvidia’s platforms like Isaac for robot simulation and Jetson for edge AI, enabling scalable Physical AI deployment.⁴ Europe’s theorists and firms, from DeepMind’s reinforcement learning to Germany’s Industry 4.0 initiatives, are well-placed to lead by combining theoretical depth with industrial scale.

Implications for Industrial Strategy

Huang’s call resonates with Europe’s strengths: a €2.5 trillion manufacturing sector, leadership in automotive robotics (e.g., Volkswagen, ABB), and regulatory frameworks like the EU AI Act fostering trustworthy AI.⁴ By prioritising Physical AI – robots that learn from human demonstration, adapt to factories, and optimise supply chains – Europe can reclaim technological sovereignty, boost productivity, and generate high-skill jobs amid the AI infrastructure surge.^2,3,4