Artificial Intelligence and its physical counterpart, Robotics, are steadily becoming part of everything we do, consider, and aspire to achieve. AI has learned to surpass human performance in tasks once believed impossible to automate. Machines can now analyze video and interpret language with levels of reliability never before achieved. Cars can operate autonomously, while intelligent systems can outperform doctors in certain medical diagnostic tasks. AI can also produce original artwork, detailed engineering designs, and promising new medicines. Yet where will this technology move next, and what limits, if any, will define its reach?
In the first part of this talk, we will take a close look at the forces truly propelling AI forward and the new forms of AI likely to emerge next. We will examine the less familiar factors behind AI’s rapid ascent, moving beyond the usual emphasis on computing power and data to consider the influence of expanding network scale, along with the growing ecosystem of intelligent agents that increasingly cooperate with other agents.
In the second part of the talk, we will explore both the current generation of AI and the future generations that have yet to arrive. We will consider the full impact of generative AI on engineering innovation and discovery, and then look further ahead to emerging forms of AI, including Physical AI in 2030 and Self-aware AI in 2040.
The talk concludes with a set of practical action items focused on preparation: how individuals can get ready for the coming transition, how organizations can adapt and respond effectively, and how national policymakers can develop thoughtful strategies for managing the broader changes ahead.
Most human activity takes place in the physical world. Yet despite its meteoric rise, artificial intelligence has remained largely confined to the virtual realm. AI can make decisions, generate ideas, analyze data, and operate complex systems. When it comes to acting in the physical world, however, robots continue to lag behind and have yet to make a substantial real-world impact.
This talk will explore the next stage of AI: embodied intelligence, where artificial intelligence moves beyond the digital domain and engages directly with the real world. It will examine why physical environments have proved so difficult for AI to master, from the uncertainty and variability of real-world conditions and the greater consequences of errors to demanding energy requirements and the scarcity of useful training data. The talk will also consider the strategies the robotics industry is adopting to overcome these barriers.
The talk will then look ahead to how robotics and other forms of physical AI may develop throughout the 2030s. It will consider the opportunities that could emerge, as well as the new technical, social, and practical challenges that may arise as AI reaches more deeply into the physical world and extends beyond its current boundaries.
Ultimately, most of human life is rooted in the physical world, and AI will not fully integrate into society until it can understand, navigate, and operate within that environment. The talk will conclude with a practical set of actions for preparing for this transition: how individuals can equip themselves for the changes ahead, how organizations can adapt and respond effectively, and how national policymakers can develop thoughtful strategies for guiding and managing the broader transformation of physical AI.