In his 2025 New Year's address, OpenAI CEO Sam Altman shifted the company's future development direction from general artificial intelligence (AGI) to "super intelligence." He expressed his confidence in building intelligent systems beyond humans in the future, and believed that this would greatly promote scientific development and social progress. The article also mentioned the limitations of current artificial intelligence technology, such as high training costs and potential errors, but he was confident in overcoming these challenges and emphasized that OpenAI is committed to the long-term interests of mankind and cautiously promotes technological development. This article outlines Altman’s vision for AGI and superintelligence, as well as thoughts on future technological security.
On January 6, 2025, OpenAI CEO Sam Altman published a New Year's speech on his personal blog, revealing that the company will begin to shift its attention from general artificial intelligence (AGI) to "superintelligence" (superintelligence). He stated in the article that OpenAI is confident in building intelligent systems that surpass humans in the future, and hopes that these systems can greatly promote scientific discovery and social development.
Altman mentioned that although current artificial intelligence technology still has some significant limitations, including high training costs and potential errors, he believes that these problems can be overcome in the near future. He emphasized that OpenAI’s mission is for the long-term benefit of mankind, and he hopes to maintain sufficient caution while advancing technological development. "We're very confident that in the next few years everyone will see the technological advancements that we've seen," he said.
OpenAI has been working hard to explore the goal of building general artificial intelligence. Altman pointed out that the definition of AGI is still vague in the industry, and OpenAI has its own view on it: that is, developing intelligent systems that are better than humans at the most economically valuable tasks. According to the agreement between OpenAI and its partner Microsoft, once AGI is implemented, Microsoft will no longer have access to its technology.
Altman specifically mentioned that the goal of superintelligence means that we can accelerate science and innovation to exceed our current capabilities. He pointed out: "With superintelligence, we can achieve everything." He has also mentioned in several previous articles that superintelligence is likely to mature in the next ten years, which will bring unprecedented wealth and prosperity to mankind. .
However, Altman also expressed concerns about the safety of superintelligence. He pointed out that humans may not be able to effectively supervise superintelligent systems, which makes people full of uncertainty about future technological development. He warned that current technical means cannot ensure reliable control of superintelligence.
Overall, Altman's New Year's post not only demonstrates OpenAI's ambitions in the field of artificial intelligence, but also triggers people's deep thinking about the security of future technology.
Altman's New Year's speech not only demonstrates OpenAI's lofty ambitions in the field of artificial intelligence, but also indicates that the development of artificial intelligence technology will face new challenges and opportunities. The exploration and concerns about superintelligence deserve the attention and discussion of the whole society to ensure that artificial intelligence technology can truly benefit mankind.