In November 2025, six preeminent figures in the AI realm—Yann LeCun, Fei-Fei Li, Jensen Huang, Geoffrey Hinton, Bill Dally, and Yoshua Bengio—converged following their joint receipt of the Queen Elizabeth Prize for Engineering. They embarked on a pivotal discussion about the future trajectory of artificial intelligence, centering on whether AI is genuinely in the throes of an industrial revolution or merely a bubble on the verge of popping.

These six laureates recounted pivotal moments of insight from their illustrious careers, moments that not only steered their research paths but also significantly shaped the course of human technological advancement. Geoffrey Hinton reminisced about his 1984 endeavor to train a miniature language model, which he deemed as the forerunner to today's large language models. Yann LeCun steadfastly maintained that the essence of intelligence resides in self-organization, rather than being dictated by instructions. Fei-Fei Li underscored the pivotal role of high-quality datasets in machine learning, highlighting how her creation of the ImageNet dataset laid the foundation for the AI visual revolution.

Bill Dally, through his innovative design of GPU computing architectures, furnished robust computational support for deep learning. Jensen Huang pointed out that AI is giving rise to entirely novel industries, necessitating the construction of AI factories worth hundreds of billions of dollars. Following the emergence of ChatGPT, Yoshua Bengio redirected his focus towards AI safety and ethics research.

When the conversation turned to whether the AI boom is a mere bubble, Jensen Huang contended that AI has birthed unprecedented new industries, far from being a bubble. Bill Dally expressed the belief that advancements in model efficiency and the exploration of applications are still in their infancy. Yann LeCun approached the topic dialectically, suggesting that while applications and infrastructure development are not bubbles, the notion of infinitely expanding along the current large model paradigm to attain human-level intelligence harbors elements of a bubble. Fei-Fei Li and Yoshua Bengio both concurred that the AI field is far from reaching maturity, with vast expanses of uncharted territory still to be explored.