Plenary: This House Believes that AI rather than Humans Holds the Key to Understanding Risk in Future.
The session took a debate format centered on the role of artificial intelligence (AI) in understanding risk, particularly in the context of environmental challenges and disaster response between two speakers: Bruno Sanchez, the executive director of an AI for Earth nonprofit, and Ivan Gayton, who leads innovation at Humanitarian Open Street Map. The central question posed is whether AI or human ingenuity holds the key to understanding risk today. Sanchez argues in favor of AI, presenting three main points. First, he asserts that society has already embraced the use of data and technology to understand risk, moving beyond the limitations of the past. He emphasizes that AI represents a significant advancement in this trajectory, likening its impact to a leap from basic arithmetic to complex mathematical transformations. This evolution allows for faster and more efficient processing of vast amounts of data, which is crucial in disaster response scenarios where time is of the essence. Sanchez’s second argument highlights the proven effectiveness of AI in delivering rapid and accurate predictions. He cites examples where computations that once took hours can now be completed in seconds, underscoring the life-saving potential of AI in emergency situations. He believes that AI not only enhances the speed of data processing but also empowers communities by providing tailored responses to their specific needs. The third point Sanchez makes is about the abundance of data available for AI applications, particularly in understanding Earth-related risks. He notes that there is a wealth of free data that can be utilized without the legal complications often associated with text-based AI. This data can be leveraged to create models that inform risk assessments and responses across various regions, from Tokyo to Colombia. In contrast, Gayton raises concerns about the implications of relying solely on AI for understanding risk. He questions who truly “holds the key” to these AI models and their predictions. He emphasizes the importance of human oversight in determining what data is prioritized and how models are trained. Gayton argues that while AI can process data and identify patterns, it lacks the creativity and adaptability inherent in human decision-making. He warns against a technocratic approach where decisions are made without community input, stressing the need for a collaborative effort between AI and human insight. The debate also touched on the limitations of AI, particularly in scenarios where unprecedented events occur, such as natural disasters that have not been previously recorded in data. Gayton points out that while AI can analyze historical data, it may struggle to predict rare but impactful events, like the Fukushima disaster, which involved a complex interplay of factors not captured in existing datasets. In conclusion, the debate encapsulates the ongoing dialogue about the role of AI in society, particularly in understanding and managing risks associated with environmental changes and disasters. It highlighted the potential benefits of AI while also acknowledging the critical need for human involvement in shaping its applications and ensuring that technology serves the broader community effectively.
“AI is actually pretty good at understanding deviations from statistical norms.” “Humans are good at understanding context, which is why they should be holding the keys.”
