|
We have been exploring a practical question: how can artificial intelligence make research outputs more accessible and useful? As AI increasingly shapes how we work, learn and make decisions, it is also beginning to transform how research knowledge is created, synthesized and communicated. Good or bad, this transformation is happening, and it raises important questions about how we adapt responsibly. Like many projects, BIONEXT produces a substantial volume of reports and datasets. Valuable knowledge is often publicly available, but the sheer volume of information can make it challenging for policymakers, practitioners and other audiences to readily interpret and use findings. This challenge inspired the development of the BIONEXT Pathways App, an AI and Large Language Model -based tool that allows users to “have a conversation” with the project. Instead of searching through dozens of documents, users can ask questions and interact directly with outputs through an interface. The app collates project knowledge and uses AI to analyse, connect and extract relevant information, making project findings more accessible and easier to use. Our hope is that this creates new opportunities for uptake and impact. If people can engage more easily with evidence and insights, BIONEXT results have greater potential to inform action and contribute to the transformative change for biodiversity. Yet perhaps the most interesting part of this work has not been the technology itself, but the conversations it has sparked around ethics and responsibility. AI can be a powerful tool, but only when applied thoughtfully and with an awareness of its limitations. We have discussed how to avoid over-reliance on AI and ensure that human judgement remains central. AI can help us explore evidence, summarise complexity and identify connections, but it should support decisions, not replace them. In developing the Pathways App, we have worked closely with colleagues developing the Multi-Criteria Decision Analysis module. Their role as “critical friends” – challenging assumptions and raising important questions – has helped us think carefully about what responsible AI use looks like, particularly in decision-support contexts. Ultimately, AI is a tool. Used carefully, it offers exciting opportunities to improve how research is shared, understood and applied, helping valuable knowledge reach the people who need it most. But the responsibility for interpretation, judgement and decision-making will always remain with us. After all, we humans are part of nature. Computers are not.
Paul Mahony is general manager at Oppla. In BIONEXT, he leads WP3 in collaboration with CzechGlobe on the topic of demonstrating transformative change in practice.
|
