Artificial General Intelligence (AGI) is the ultimate goal in the AI arms race, with many of the world’s top researchers and innovators working toward creating machines that can think, reason, and solve problems as humans do.
While AGI remains a theoretical concept, experts disagree on how close we are to reaching it. Some believe AGI could emerge within the next few years, while others argue that it may take decades.
What is AGI?
Before delving into expert opinions, it’s important to understand what AGI is.
Unlike the narrow AI systems we have today, which are designed to perform specific tasks (such as language translation or image recognition), AGI would be capable of performing any intellectual task that a human can do.
It would have the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly, and learn from experience.
Demis Hassabis: A Vision for AGI in the Next 5-10 Years
Demis Hassabis, the CEO of Google DeepMind and recent Nobel laureate, is one of the leading voices in the push toward AGI.
According to Hassabis, we can expect AGI to emerge in the next five to ten years.
During an appearance on 60 Minutes in April 2025, he suggested that by 2030, AGI will be integrated into daily life, with systems capable of deeply understanding and interacting with the world in nuanced ways.
Hassabis acknowledged that current AI systems lack true curiosity, imagination, and intuition, which are key aspects of human intelligence.
He believes that within the next decade, we will see systems that can not only solve important scientific problems but also come up with these problems in the first place.
Sam Altman: AGI Will Arrive Soon
Sam Altman, the CEO of OpenAI, is another AI leader who believes that AGI is closer than many think.
Altman has expressed excitement about the potential arrival of AGI in 2025, predicting major strides toward this goal in the near future.
His optimism stems from the rapid progress made by OpenAI and other organizations working on advanced AI systems, suggesting that the capabilities of current models are advancing rapidly.
Miles Brundage: AGI in the Coming Years
Miles Brundage, former head of AGI readiness at OpenAI, predicts that AGI will manifest in the next few years.
According to Brundage, AI systems will soon reach a level where they can perform virtually any task a human can do remotely, including operating devices like a mouse or keyboard, or even engaging in video chats.
This suggests that we might see specialized forms of AGI that can carry out tasks that were previously thought to be impossible for machines.
Dario Amodei: AGI by 2026
Dario Amodei, the CEO of Anthropic and former research lead at OpenAI, is another expert confident in the imminent arrival of AGI.
He predicts that by 2026, we will see powerful AI systems that outperform human intelligence across a range of fields.
These systems will be multimodal, capable of performing multiple types of tasks, and will be highly independent and fast. Amodei compares this future AGI to “a country of geniuses in a data center,” reflecting the immense power and capability of such systems.
2026Geoffrey Hinton: A Wide Range of Possibilities
Geoffrey Hinton, often referred to as the “godfather” of AI, has offered a more flexible timeline for the arrival of AGI.
In a post on X (formerly Twitter), Hinton predicted that we might see AI that surpasses human intelligence in as little as five years, but he acknowledged the uncertainty of such predictions.
He placed the range for AGI’s arrival between 5 and 20 years, reflecting the unpredictable nature of AI development.
Andrew Ng: Cautiously Optimistic
Andrew Ng, a renowned AI researcher and co-founder of Google Brain, takes a more conservative stance on AGI.
While Ng believes that AGI could eventually perform tasks as complex as writing a Ph.D. thesis or flying a plane, he is not sure if it will happen soon.
Ng encourages skepticism toward companies that claim AGI is imminent and emphasizes that achieving AGI should be considered a long-term goal, with caution taken to ensure its responsible development.
Richard Socher: AGI in Three to Five Years?
Richard Socher, former chief scientist at Salesforce and now CEO of You.com, offers two potential timelines for AGI.
The first is an economic definition of AGI, where 80% of jobs are automated by AI. He predicts this could happen within three to five years.
The second, more ambitious definition of AGI, would involve a machine learning system with the cognitive abilities of a human, capable of learning in a similar way.
For this level of AGI, Socher suggests that it could take anywhere from 10 years to 200 years.
Yann LeCun: AGI Is Still Far Off
Yann LeCun, Meta’s chief AI scientist and another prominent figure in the AI field, is less optimistic about the immediate future of AGI.
During an appearance at the World Economic Forum in January 2025, LeCun stated that AGI is not “around the corner” and that it could take decades before we see any meaningful breakthroughs.
He also criticized the popular notion, perpetuated by science fiction, that AGI will suddenly “appear” as a singular event.
LeCun believes that the development of AGI will be gradual, with small incremental advances leading to significant progress over time.
Conclusion: How Far Are We from AGI?
The question of when AGI will arrive remains highly debated among experts. While some, like Demis Hassabis and Sam Altman, predict that we are on the cusp of achieving AGI, others like Yann LeCun suggest it could take decades of gradual progress before AGI becomes a reality.
The range of predictions from these AI leaders reflects the uncertainty and complexity of the journey toward AGI.
While we may not have a clear answer yet, one thing is certain: the development of AGI will be a transformative event in human history, with the potential to reshape industries, economies, and even society itself.
As research and development continue at a rapid pace, the dream of AGI is becoming ever more tangible—and with it, new possibilities and challenges for the future.