*Subtitles: Choose “Subtitles/CC” in the “Settings” (gear icon) at the bottom right of the YouTube video.
*Subtitles may not show with multiple languages or overlapping audio.


The data on biodiversity is enormous, and what we know today is only a fraction of it. Biodiversity data is complex and has been limited in its ability to be collected on a large scale. In the world of 2050, to what extent will it be possible to collect such information, decipher it, and take action based on it? How far should we be able to collect such information, decipher it, and take action based on it in the world of 2050? We will hear frank opinions from young leaders of the world, including CEO of an AI startup, a biologist, and a nature conservationist.

【Reflection】
This session featured in-depth discussions on how to implement Nature Positive by 2050 using AI and big data, drawing on diverse field knowledge and perspectives from research and business. The moderator first guided the audience and speakers through a “visualizing 2050 concretely” exercise, encouraging them to imagine details like colors, shapes, and smells before returning to the present reality of 2025. This established a clear stance: even when discussing technology, the relationship between humans and nature remained central. This introduction established a shared premise: data and AI can serve not as tools for top-down management of nature, but as mediators supporting symbiosis rooted in human sensibilities and ethics.

The first point of discussion was clarifying the utility and limitations of big data/AI. Wakana Inoue, drawing from mangrove conservation practices in Fiji and Kiribati, emphasized that while satellite data is effective for assessing the health and growth potential of planted forests across wide areas, it should not replace local experiential knowledge but rather support it. For example, the approach should respect residents' intuitive judgment about “where to plant” and use data to reinforce that. She also introduced an initiative where sales from crayons made from fallen leaves are reinvested in reforestation, noting that visualizing the joy of participation itself can become a driving force for sustainability.

Shunsuke Tsuboi presented concrete examples of optimizing soil, crop planting, and fertilization using satellite data combined with AI, while stressing the importance of translating this into actionable forms for field use. He explained that in developing regions, where purchasing power is constrained, companies and governments must bear the costs and design schemes that deliver incentives to farmers. He also mentioned low-barrier implementation scenarios, such as AI providing individual farmers with fertilization and cultivation advice simply by capturing audio or photos during meetings or gatherings. Regarding the distance created by technology's energy demands and abstract nature, he stressed that design must convey value to users. Responding to the moderator's question, he repeatedly highlighted the dual pillars of “access” and “incentives.”

Anna Reyes, drawing on conservation experience in the Masungi Geopark in the Philippines, raised the importance of data sovereignty and equitable governance. She noted that AI and sustainability frameworks often rely on Western-centric norms, risking the exclusion of cultural contexts from the Global South. She argued for a bottom-up framework that curbs the arbitrariness of corporate reporting through third-party verification while avoiding top-down “false solutions.” Furthermore, he urged audience engagement with the message that “what is unknown and unmeasured is undervalued,” emphasizing that “measuring” is the starting point for fostering affection and protection.

Wildlife expert Seung-Yun Baek introduced practical applications using GPS tracking, camera traps, and AI image recognition to understand animal movements and food resource availability, enabling the prediction of human-wildlife conflict risks in advance. However, he emphasized that while the causes of conflicts are largely understood, society's failure to act remains the core issue. He stressed the need for mechanisms linking data value to tangible action.

Midway through the discussion, the moderator deliberately raised the “danger of viewing the world solely through a technological lens.” In response, the panelists carefully addressed both the potential of AI (visualizing patterns, early warning, efficient resource use, etc.) and the requirements for its “use” to amplify connections with nature and local communities. Inoue repeatedly stated, “Data is support, not replacement.” Tsuboi touched on a future vision where employment and on-the-ground connections in rural areas increase, positioning AI as a tool that encourages humans to return to the field.

During the Q&A, attendees posed sharp questions about decolonizing data. Concerns centered on how to safeguard decision-making authority and prevent the erosion of agency and discretion for Global South stakeholders within structures where data infrastructure and AI are predominantly designed by English-speaking, developed nations.Anna Reyes acknowledged that many existing frameworks rely on Western norms, while reiterating the need to transition towards locally-driven frameworks. Through the exchange with the audience, a deeper understanding emerged that the key to implementation lies in balancing governance—including ownership, verifiability, and accountability—with the visualization of emotions like joy and pride.

Through the discussion, speakers indicated a direction where the use of big data/AI should be designed as a tool supporting the “co-education” of people and nature, by creating a back-and-forth between measuring (quantification) and sensing (relationships/narratives). The discussion also highlighted risks such as the energy demands of technology, privacy concerns, and regional disparities. It was shared that we must refine effective “uses” for 2050 through incentive design and bottom-up consensus building.

【Post EXPO Initiatives】
The speakers highlighted the following as key perspectives for their respective research and business development after the conference:

(1) Strengthening efforts to provide data and technology in ways that are meaningful for local communities.
・Develop accessible dashboards and actionable metrics aligned with local priorities and traditional knowledge, connecting them to decision-making by residents, field practitioners, and policymakers
・Advance responsible and equitable data governance frameworks, embedding data sovereignty, fair benefit distribution, third-party verification, and accountability into operational rules.
・Foster collaboration among businesses, startups, governments, and regions to clarify incentive structures, creating an environment where optimized practices can be adopted without overburdening agriculture or conservation sites.

(2) Research and Technology Development
・Integrate data from satellites, sensors, camera traps, GPS, etc., using AI to advance monitoring that enhances ecosystem understanding and protection while minimizing the system's own environmental impact
・In urban areas, expand models that incorporate biodiversity value and ecosystem services (flood mitigation, heat island reduction, physical and mental health) into design and operational KPIs alongside aesthetic considerations for green spaces, enabling evaluation from the planning stage
・In education and outreach, implement programs fostering wonder and emotional connection to nature, particularly among younger generations. Provide accessible information to shift consumer and voter choices toward biodiversity-conscious options. Through these efforts, strengthen collaboration among stakeholders and implement nature-positive outcomes in society.

*Part or all of this report was generated by AI.