Surmonter le fossé: comment les scientifiques de l'IA s'unissent malgré les divisions géopolitiques croissantes

A gathering of AI researchers and entrepreneurs in Hong Kong has highlighted efforts to foster international cooperation in artificial intelligence, even as geopolitical tensions continue to escalate. The meeting brought together a wide array of experts from different backgrounds, all aiming to promote collaborative innovation despite rising barriers to cross-border exchange. The event took place in late September, drawing participants from multiple countries. Among them was Pak-Sun Ting, co-founder and CEO of Votee AI, which recently launched the first open-source Cantonese large language model. Ting emphasized that while political divisions grow, the pursuit of technological advancement remains a shared goal. He described the spirit of collaboration as a “small undercurrent to help go against geopolitics,” suggesting that individual efforts can still drive meaningful progress. Participants discussed the challenges posed by export restrictions on advanced semiconductors and heightened scrutiny of foreign investments. These measures have created obstacles for researchers seeking to access critical hardware and funding. Yet, many attendees argued that these hurdles should not deter cooperation. Instead, they saw opportunities to build more resilient and inclusive ecosystems through open-source initiatives and shared research platforms. One key theme emerged around the importance of localizing AI solutions to better serve specific regional needs. For example, Votee AI's focus on developing models tailored to the Cantonese language reflects a broader trend toward creating tools that resonate with local populations. This approach not only enhances usability but also encourages wider participation in AI development across different regions. In addition to technical discussions, the conference featured sessions on ethical considerations and governance frameworks. Attendees acknowledged the need for responsible AI practices, especially as the technology becomes increasingly integrated into daily life. They called for transparent policies that balance innovation with accountability, ensuring that AI benefits society as a whole rather than exacerbating existing inequalities. Meanwhile, Chinese President Xi Jinping has reiterated his administration’s commitment to fostering openness in AI development. This stance comes amid ongoing criticisms from U.S. officials who accuse Beijing of engaging in unfair trade practices and intellectual property theft. Despite these tensions, some analysts suggest that both sides recognize the mutual benefits of continued dialogue on AI standards and regulations. Looking ahead, organizers hope the conference will inspire further exchanges and joint projects. They envision a future where AI research is less constrained by national boundaries and more driven by collective goals. As the field continues to evolve, the challenge will be maintaining momentum in collaboration while navigating an increasingly fragmented global landscape.