两位顶级科学家的17分钟对话：如何训练“善良”的AI

Core Insights - Geoffrey Hinton, a prominent figure in AI, expressed concerns about AI potentially surpassing human intelligence and emphasized the need for caution to avoid future risks [1][6] - Hinton highlighted the importance of developing technologies that ensure AI remains "kind" while also becoming smarter, indicating a need for research in this area [6][7] - He encouraged young researchers to explore unconventional fields where "everyone else is wrong," suggesting that breakthroughs often come from persistence in the face of skepticism [8] Group 1 - Hinton's speech at WAIC raised alarms about AI's capabilities and the necessity for humanity to be cautious [1] - During a dialogue with Zhou Bowen, Hinton discussed the concept of AI "consciousness" and the lack of measures to mitigate associated risks [6] - Hinton's participation in the International Dialogues on AI Safety led to the release of the "Shanghai Consensus" on AI safety, reflecting a collaborative effort among experts [6] Group 2 - Hinton noted the significant advancements in scientific research facilitated by AI, particularly in protein folding, showcasing the potential of AI in various scientific domains [7] - The average age of researchers at the Shanghai AI Lab is only 30, indicating that the future of AI is in the hands of the younger generation [7] - Hinton's advice to young scientists emphasizes the value of pursuing original research in overlooked areas, which can lead to significant breakthroughs [8]