Recently, OpenAI CEO Sam Altman clearly denied the claim that "OpenAI has achieved general artificial intelligence (AGI)" on social platform X. He said that despite the intensifying discussion about AGI, OpenAI will not deploy AGI next month and has not developed this technology. Ultraman reminds everyone that they should lower their expectations "100 times."
AGI, or general artificial intelligence, refers to an artificial intelligence that can understand, learn and perform any human intelligence task. Recently, some AI fans have mistakenly interpreted some content posted by OpenAI employees as evidence that AGI has been implemented, leading to numerous misunderstandings. Although OpenAI is indeed working hard to launch a series of new AI inference models and released some related updates last month, the realization of AGI is still far away.
For some time, Altman and the OpenAI team have been actively promoting new inference models. Altman said in a podcast that he believes OpenAI has mastered the key technologies to build AGI, and predicted that this year may see AI agents enter the workforce for the first time. He also mentioned that within the next few thousand days, humans may see the emergence of super intelligence.
Recently, however, some commentators, notably AI-focused author Gwern Branwen, have once again fueled rumors that OpenAI may implement AGI in-house. In a post on social media, he said that OpenAI is on the verge of a breakthrough, and even speculated that it may have crossed a critical tipping point. In this regard, Noam Brown, a top researcher at OpenAI, also spoke out, pointing out that there is a lot of vague AI hype on social media. Although he is optimistic about further progress, there are still many unsolved research problems that need to be overcome.
Although OpenAI continues to make progress in AI technology, it still has a long way to go before truly realizing AGI. Regarding the expectations of fans, Ultraman's appeal is particularly important.