Elon Musk recently tweeted that he thinks Grok 5 has a chance of reaching AGI, meaning that we might see it a lot earlier than all previous predictions have suggested – perhaps as soon as the end of this year, when Grok 5 is set to be released.
“I now think @xAI has a chance of reaching AGI with @Grok 5,” tweeted Musk on his X.com platform. “Never thought that before.”
I now think @xAI has a chance of reaching AGI with @Grok 5. Never thought that before. https://t.co/FaBUYegl3DSeptember 17, 2025
Artificial general intelligence (AGI) is the goal that all AI companies are striving to achieve. While definitions of AGI keep shifting subtly to conveniently meet the requirements of whatever point people are trying to make, most experts agree that, essentially, it’s the idea of an artificial intelligence system that can think, learn, and apply knowledge across a very wide range of tasks to the point where it has the same flexibility and adaptability that humans have.
AI companies are striving for AGI because once reached, it opens up dramatic possibilities for breakthroughs in science, medicine, technology, and our daily lives in general. But it also raises huge ethical, safety, and control challenges, because we wouldn’t necessarily be able to control something that can out-think us. We’ve all watched the Terminator movies, after all.
The AI Godfather
Geoffrey Hinton, a British-Canadian computer scientist who pioneered deep learning and is often called the ‘Godfather of AI,’ has been particularly vocal in raising concerns about the potential impact of developing AGI, estimating a “10-to-20% chance” that AI might lead to human extinction within the next 30 years.
The tweet by Musk is something of a turnaround, as he’s previously been quite a vocal critic of AGI in general. Just two years ago, Musk was a signatory to a global appeal advocating for a moratorium on advanced AI experiments, citing “profound risks to society and humanity” and equating AGI with something “significantly higher risk than nuclear weapons.”
He further advised against a headlong rush into its development without due consideration of the potential consequences. If Musk is now saying that Grok 5 will usher in the AGI era, shouldn’t we be worried? Especially if that power is held by a single company controlled by an idiosyncratic individual like Musk.
Maybe it’s too soon to start worrying
To find out I talked to Himanshu Tyagi, a professor at the Indian Institute of Science and co-founder of Sentient, the $1.2bn open-source AI unicorn backed by Peter Thiel. According to Tyagi, we might have nothing to worry about just yet.
“We are seeing extraordinary improvement in AI’s ability to handle complex digital tasks,” Tyagi told me. “We should expect Grok 5 to be able to conduct complex research over the internet and give extraordinary answers. You can call that AGI. But will it solve new scientific problems, discover new synthetic proteins, anytime soon? I doubt it.”
So, while advanced AIs like Grok 5 or the next version of ChatGPT may dominate digital tasks, it doesn’t sound like full AGI will be coming quite yet.
That doesn’t mean an AGI-light version will be without consequence, even if this isn’t the full AGI we were originally promised. Opaque AI models from companies like xAI and OpenAI often seem to prioritize speed over safety, potentially amplifying existential threats like like security and safety, and open source alternatives like Sentient offer a different vision of the future, one which still remains unwritten.
Add Comment