According to a recent post by OpenAI CEO Sam Altman, the company has figured out how to build artificial general intelligence (AGI) and is now shifting its focus to superintelligence. Based on reports from TechCrunch, Altman believes superintelligent tools could accelerate scientific discoveries and innovations beyond human capabilities.

Altman previously mentioned that superintelligence might be “a few thousand days” away. He also noted that its arrival will be “more intense than people think.” In his post, Altman explained that AI systems that outperform humans at most economically valuable work would be considered superintelligent. OpenAI and its partner Microsoft have defined AGI as systems that could generate at least $100 billion in profits.

Although Altman did not clarify the exact definition of superintelligence he is referencing, it seems likely that he refers to AI systems capable of performing tasks autonomously. Based on reports from TechCrunch, Altman suggested that AI agents might “join the workforce” and potentially change how companies operate in the coming year.

However, today’s AI systems still face technical limitations. These include hallucinations, making obvious mistakes, and high operational costs. Despite these hurdles, Altman remains confident that OpenAI can overcome these challenges quickly.

Altman also expressed his belief that in the next few years, the world will recognize the potential of superintelligence. According to Altman, OpenAI aims to maximize the broad benefits of AI while acting with great care.

OpenAI has acknowledged the risks of superintelligent AI in previous reports. As stated in a 2023 post, the company admitted it does not yet have a solution for steering or controlling superintelligent systems. OpenAI has also expressed concerns about the difficulty of reliably supervising AI systems that are far smarter than humans.

Recently, OpenAI disbanded its AI safety teams, including those focused on superintelligent systems. Several key safety researchers left, citing OpenAI’s increasing commercial focus as a reason for their departure.

Altman responded to critics who claimed OpenAI wasn’t focusing enough on safety by pointing to the company’s track record.

SOURCE: TECHCRUNCH | PHOTO: SEONGJOON/EXPANSION