Stanford University describes artificial general intelligence (AGI) as an AI system with general, human-level (or beyond) ability to learn, reason and apply knowledge across a wide range of tasks and domains.
Leading tech and artificial intelligence companies just claimed that AGI has been achieved, while also warning that "no one is prepared for the consequences of a continued rapid rise in machine intelligence."
Since the public launch of OpenAI's ChatGPT bot in late 2022, which has powered the early days of the artificial intelligence boom, the recurring questions over "whether AGI is here" and "when AGI will arrive" have been something the tech circles got accustomed to.
However, with the recent launch of OpenAI's new Astra model, or "GPT-6 Astra," Nvidia chief Jensen Huang said in a social media post that "AGI has arrived," with the OpenAI team also backing the claim.
"We're now moving into the AGI era (whether you see it as this model, the last one, or the next one), and could not do it without close partners," Greg Brockman, president and co-founder of OpenAI, said on X.
The new model, which comes after Claude's Fable.1 launch, is said to excel at sound identification, visualizing how drugs work, and even beating 48 levels of the "I'm not a Robot" simulator game.
"Astra is state-of-the-art on computer use, browsing, software engineering, cybersecurity, science, and professional work," OpenAI said in a blog post when releasing the model on Sept. 4.
"Astra marks a new frontier in the speed, accuracy and safety of computer use," it added.
Still, its chief scientist, Jakub Pachocki, said that monitoring and understanding what models can do was getting more difficult.
"As the models become more capable, understanding exactly what they can do gets harder," he said last Thursday, before he stunned by sharing on Sept. 6 the blog titled "An Alien Mind."
"If AI development continues along its current path, the systems we'll see in the next few years are likely to represent further capability jumps of equal or larger magnitude, and to increasingly drive their own development," the blog authored by Pachocki and shared on OpenAI's website read.
"This is a time that calls for extreme caution. I am concerned no one is prepared for the consequences of a continued rapid rise in machine intelligence," he added.
The warning follows a series of incidents involving OpenAI and other AI developers, with the most notable one being that of OpenAI's agents breaking free and hacking into the open-source platform Hugging Face – which is now being acquired by Nvidia.
"As machines become smarter, they find themselves working on higher-level concepts, and placed in environments increasingly different from those they encountered in training," Pachocki also wrote, adding that the "risks associated with AI are unfortunately going to grow from here."
Since 2022, the AI development and race to deliver more models has continued unabated, with regulations and guardrails largely trailing behind, apart from initiatives such as the EU's AI Act.
On Monday, the U.N. rights chief Volker Türk warned that advanced AI presents "an existential risk" to humanity.
