Recently, the CEO of Anthropic, Dario Amodei, published a signed article, publicly calling on the entire artificial intelligence industry to slow down the pace of technological iteration.
In his article, Amodei made a sharp core judgment: if humans do not actively constrain the development rhythm of AI, the speed of future technological breakthroughs will soon completely exceed the understanding of ordinary people and completely go beyond human control.
Soon after Amodei's article was published, Elon Musk publicly expressed his support for Amodei's view; Sam Altman, CEO of OpenAI, also quickly responded publicly, clearly stating that he fully agreed with the core idea that "the frontier of AI does indeed need to appropriately slow down its development pace."
The Most Difficult Practical Dilemma: Who Should Slow Down First
However, when discussing the practical feasibility of this global joint slowdown plan, Amodei frankly mentioned that China is the most difficult practical dilemma in his view during the process of advancing the plan.
He directly pointed out the core contradiction of this concept: if only some countries and a few leading technology companies choose to slow down and reduce their R&D pace at present, while other entities are not restricted and continue to push forward with full force, the parties that voluntarily slow down will inevitably fall into a very disadvantageous position in terms of technical competition discourse, industrial economic benefits, and even future military technology advantages.
Amodei believes that from a long-term perspective, the international community should try to establish a unified mechanism for controlling the development speed of artificial intelligence through multilateral cooperation, to avoid uncontrolled competition pushing technological capabilities rapidly beyond existing safety boundaries. However, he also acknowledged that the implementation difficulty of such transnational cooperation is extremely high - because the strategic benefits of being the first to master the next generation of advanced artificial intelligence are too great, and it is almost impossible for any entity to give up the opportunity to lead in this key technology track.
He admitted that he still hasn't been able to determine whether this globally coordinated mechanism can truly take effect, but due to concerns about the long-term risk of technological loss of control, he still believes that the entire industry has a necessity to make efforts in this direction.
