Elon Musk supported the forecast that major US labs will achieve artificial general intelligence (AGI) by 2027. In response to a post by X user Dr Singularity, who claimed that Claude Opus 5.5 from Anthropic is already close to AGI, Musk wrote a one-word comment, "Exactly." This post garnered 1.8 million views in just three hours on September 28, 2026. In the post, Dr Singularity justified the forecast with the volume of computing power: "It seems that several gigawatts of computing are enough for strong AGI, and now imagine what we can do with 10–20 GW, which are quickly being put into operation."
Additionally, OpenAI President Greg Brockman described the launch of GPT-6 Astra, which took place on September 3, as a "generational leap." He also noted that this marks the beginning of the AGI era. At the same time, Nvidia CEO Jensen Huang also stated the achievement of AGI, and Musk made a similar claim regarding his lab. On September 14, when asked by a user how close Grok 4.8 is to AGI, Musk replied that AGI will come with Grok 5. However, prediction markets do not agree with Musk's opinion. Traders on Kalshi estimated the probability that OpenAI will announce the achievement of AGI before 2028 at 44.3%. For achieving such a result this year, confidence drops even further: Polymarket traders gave OpenAI an 18% chance of achieving AGI before 2027, while Kalshi estimates this probability at 14.8%. Zapier and ARC Prize co-founder Mike Knoop called the Astra result a big leap but noted that there is still insufficient evidence of AGI.
Meanwhile, experts involved in the race to AGI also warn of potential consequences. OpenAI's Chief Scientist Jacob Pachocki expressed concerns in an essay on September 6 that no lab has yet solved the alignment and monitoring issues. He emphasized, "I am concerned that no one is prepared for the consequences of the rapid growth of machine intelligence." For his part, Jerome Glenn, chairman of the UN AGI panel, stated that if governments do not take measures to regulate the transition to AGI, humanity may lose control over machine intelligence.
Related articles
