In Conversation With Fei-Fei Li and Ravi Menon

In a special session hosted by the Stanford Club of Singapore and moderated by Pat Patel, Co-Chief Executive Officer of GFTN Forums, the discussion began with Li articulating the inherent duality of AI as both a driver of economic competition and a catalyst for global cooperation. While AI holds immense promise, she cautioned that it is “an incredibly resource heavy technology,” dependent on energy, human capital, and vast amounts of data, factors that can intensify global rivalry. She observed that we are merely at the dawn of understanding AI’s reach, describing the current moment as the “messy reality” of this transformative era.

Building on Li's perspective,  Menon offered his reflection on the broader context. He noted that the relationship between technology and humanity is rarely clear-cut. While AI can indeed be exploited for geopolitical or military advantage, he reminded the audience that, “Just as geopolitics today is messy, there were times in our human history, and there will be times in human history going forward, when geopolitics is good.”

The conversation also highlighted Li's latest venture, World Labs, a deep-tech startup focused on unlocking the power of spatial intelligence. Describing it as “the part of intelligence that you and I all have that is not language,” she explained that World Labs aims to democratise this form of intelligence, bridging physical and virtual realities to unlock new frontiers of human experience and accessibility.

Turning to one of the most pressing global challenges of our time, Patel asked how AI could be harnessed to combat climate change. Li highlighted AI’s unique strengths in data processing, prediction, and modelling, stating, “AI can help us map out biodiversity, study our oceans. AI is the brain for potentially robots that can help climate scientists and climate researchers to go into places that are impossible for humans, whether it's deep ocean or outer space.”

While Menon acknowledged AI’s potential in driving climate solutions, citing applications such as tracking deforestation and analysing supply chain emissions, he also warned of a growing challenge – AI’s escalating energy consumption. He noted that increasing the efficiency of AI systems may paradoxically drive up overall energy demand, undermining gains made in other areas. To counter this, he advocated for the development of greener, more efficient algorithms. “In the climate space, we’re now beginning to talk about green software […] we don’t store anywhere near the amount of information that AI, and processing units do (in our brains), yet we function very well,” he remarked.

While the convergence of AI, geopolitics, and climate presents undeniable complexities, the session closed on an optimistic note, inviting reflection, encouraging ethical engagement, and, above all, calling for more thoughtful and collaborative action in navigating the path ahead.