The question of when artificial intelligence will match human intelligence remains one of the most contentious debates in technology. At the recent FT Future of AI Summit in London, the world's leading AI pioneers, all recipients of the 2025 Queen Elizabeth Prize for Engineering, revealed dramatically different predictions, highlighting a fundamental split in the scientific community.
A Spectrum of Predictions on AI's Future
The panel, featuring the architects of modern AI, showcased a wide range of viewpoints. Geoffrey Hinton, often called the Godfather of AI, offered one of the most optimistic timelines. He stated that if artificial general intelligence (AGI) is defined as a machine that can always win a debate, we are likely to see it within the next 20 years. Hinton traced this progress back to his 1984 work on a tiny language model, noting that the 40-year journey was stalled only by a lack of computational power and data, hurdles that have now been overcome.
In contrast, Yoshua Bengio presented a more nuanced outlook. He pointed to the exponential growth in AI's planning capabilities over the last six years, suggesting that AI could reach the competency level of an employee in their job within about five years. However, he was quick to add a note of caution, emphasizing the inherent uncertainty in such forecasts and advising against making definitive claims.
Offering the most subdued perspective was Meta's chief AI scientist, Yann LeCun. He argued that the arrival of human-level machine intelligence will not be a single event but a gradual expansion of capabilities across different domains over the next five to ten years. LeCun highlighted a critical scientific gap, noting that current AI systems are not as intelligent as a cat, and that future progress depends on new paradigms, not just more data and infrastructure.
Is the Question Itself the Problem?
Not all experts agreed that matching human intelligence is the right goal. NVIDIA CEO Jensen Huang dismissed the question as largely academic. He argued that the practical application of ever-improving AI technology to solve important problems is what truly matters, making the timeline for human-level intelligence irrelevant.
This sentiment was echoed by Bill Dally, NVIDIA's chief scientist, who clarified that the primary goal of AI should be to augment and complement human abilities, not to replace humans. He stated that the objective is to free up humans to focus on uniquely human traits like creativity, empathy, and social interaction.
Stanford Professor Fei-Fei Li provided a crucial perspective, explaining that machine and human intelligence are built for different purposes. She noted that AI already surpasses human capability in specific tasks, such as recognizing thousands of objects or translating between 100 languages. However, she highlighted a significant frontier where AI still fails: spatial intelligence. This underscores that human intelligence is a complex blend of abilities that extends far beyond language.
AI Boom: Sustainable Revolution or a Bubble?
The panel also addressed concerns about whether the current AI boom is a sustainable revolution or an inflated bubble. Jensen Huang drew a sharp contrast with the dotcom era, stating that today, almost every GPU is actively being used, unlike the dark fiber deployed during the internet bubble. He justified the massive infrastructure investments, stating that AI represents a fundamental shift as it addresses labor and work itself.
Yann LeCun, however, identified a potential bubble within the broader revolution. He expressed skepticism that the current paradigm of large language models (LLMs) alone can be pushed to achieve human-level intelligence, a belief he does not share.
Despite the disagreements on timelines and definitions, the pioneers were unanimous on one point: the transformative power of AI. As Fei-Fei Li concluded, AI is a "civilisational technology" that will impact every human and every business sector. The debate on when AI will match human intelligence may never have a single answer, but its profound impact on our world is already undeniable.
