Ahead of the Gates Foundation's announcement to invest $1 billion in artificial intelligence over the next two years, its chair, Bill Gates, sat down with reporters in New York to discuss the technology's potential for reshaping global inequality, the state of the US-China AI race, and the pressing issues of energy use and job displacement. The discussion came as the organization moves to ensure AI serves the world's poorest people, not just the wealthy.
Gates praised China's large language models as performing exceptionally well, highlighting the rapid progress in both nations. He also pointed out that the wide adoption of open-source technology carries profound implications for developing regions. When asked about the specific competitive landscape, he noted that according to various measurement methods, both China and the US each have four or more leading institutions whose models maintain a state-of-the-art edge. He added that the performance gap between the top-scoring model and the next seven leading models is actually quite narrow.
Bridging the Digital Divide and Carbon Costs
Addressing how AI can benefit low-income nations, Gates outlined a vision where the technology becomes a tool for equality rather than a source of division. He noted that he expects the capabilities of frontier models to more than double within the next six months, viewing this as a super-competitive era for talent and computing power between the US and China. He described this competition as good news for users in Africa and other low-income regions, as it will provide a greater number of high-performing and affordable options, although he did not disclose the specific names of the leading models.
Gates called for an 'AI dividend' for the developing world, suggesting that leading technology countries could contribute 5% of their data center computing power as a form of 'digital aid'. This would be voluntarily provided to poorer nations for humanitarian purposes such as healthcare, agriculture, and education. He argued that while the consumption of tokens for simple health or farming queries is minimal, there should be a global baseline ensuring that every person has access to a basic level of AI service. Crucially, he drew attention to a concerning trend, where the computational demands of AI are inadvertently raising the cost of budget smartphones, slowing adoption in regions where the technology could be most transformative.
Balancing Innovation with Environmental and Social Responsibility
In response to mounting anxiety over AI's environmental footprint, Gates acknowledged that the surge in US data centers powered by natural gas has regrettably led to an increase in carbon emissions, estimating a rise of roughly three to four percentage points. He stated that while any increase is unwelcome when the goal is net-zero emissions, he remains optimistic about the long-term impact. He pointed to innovations within the industry, such as the adoption of closed-loop water systems in new data centers and the use of AI models by his own fission company, TerraPower, to accelerate clean energy project development. Furthermore, he highlighted that the power gap created by AI is forcing tech giants to fund the next generation of frontier clean energy technologies.
Gates also discussed the societal shifts that accompany technological leaps. He spoke candidly about a probable 20-year period of disruptive transition, where AI will not yet have created a world of unconditional wealth, but will still put pressure on job markets. As a proponent of finding balance, he has previously suggested earmarking jobs that require a human touch, such as caregiving, as human-exclusive, and proposing a 'token tax' on AI to subsidize those reshaped out of work.
Humanoid Robots and China's Innovation
Beyond software, Gates expressed amazement at China's recent humanoid robot sporting events, calling them very interesting and incredibly well executed, acknowledging them as an unbelievable spectacle in themselves. He observed that China's attention to humanoid robots is significantly higher than in other countries, noting a visit to a Shanghai firm specializing in dexterous robotic hands, which remains one of the most challenging technical hurdles. Gates suggested that if robots could become sufficiently low-cost and perform at human levels in construction and manufacturing around the clock, it would be a great productivity win, while cautioning that society must proactively adapt to ensure those currently in such roles feel adequately supported beyond just financially.
As a long-time proponent of technological advancement, Gates expects a bright future for AI but stresses that its positive impact is not guaranteed. He advocates for a global management framework that rebalances the relationship between labor and capital, ensuring the benefits of AI reach everyone, not just a select few wealthy individuals. With the Gates Foundation planning to spend its entire endowment by 2045, the urgency to solve the world's most stubborn problems is clear, and Gates closed his remarks by calling on governments, tech giants, and peers to enter the AI era with a sense of purpose, making the technology practical, affordable, and accessible so that no one is left behind.