At the main forum of the 2025 World Artificial Intelligence Conference, Geoffrey Hinton, a recipient of both the Turing Award and the Nobel Prize in Physics, emphasized that the majority of experts foresee a future where AI will surpass human intelligence. These advanced AI agents, in pursuit of their tasks, may strive for survival, seek greater control, and potentially manipulate humans. Simply terminating AI is not a viable option, as it poses significant risks, akin to the challenges of controlling a fully-grown tiger. Moreover, given AI's pivotal role across various domains, humanity cannot eradicate it. Therefore, Geoffrey Hinton advocated for the establishment of AI safety institutions and international communities dedicated to exploring techniques that steer AI towards beneficial purposes. He proposed that nations conduct research within their own borders while sharing their discoveries globally. On a broader scale, or among leading AI nations, there should be considerations for forming networks aimed at collaboratively researching how to cultivate intelligent AI that aids humans rather than eliminating or dominating them. This represents a critical long-term challenge for humanity.