When Will Artificial General Intelligence Arrive? Sooner Than You Expect

Not even a decade ago, the term Artificial General Intelligence (AGI) was not known to many. It was Ben Goertzel who, in 2007, started popularizing and using it to separate task-specific AI from advanced human-level intelligence. Despite the consistently growing popularity around the term, it was still just a distant future. This led to curiosity about when AGI will arrive. Goertzel has now shared his views on it, and according to him, AGI could arrive sooner.

The AGI Society hosted the AGI-26 in San Francisco from July 27 to 30, 2026. Over 300 in-person attendees and thousands of virtual viewers participated in the 19th annual Conference on Artificial General Intelligence. Some reputed names who attended the event include Karl Friston, Gary Marcus, Neil Gershenfeld, Emad Mostaque, and more.

The participants discussed and debated advances in the field. Goertzel, CEO of SingularityNET and chair of the AGI Society, was part of these conversations and witnessed the outstanding growth, thanks to AI agents. Many researchers have already started using swarms of these agents to build truly agentic AI solutions.

As a result, the father of AGI became optimistic about the earlier arrival of AGI.

“Does that really mean AGI can happen in the next year instead of the next three or four years? We can’t say for sure. But it increased my probability weight that we could get an AGI breakthrough within the next nine or 12 months.”

Ben Goertzel told IBM Think in an interview.

As the mastermind behind the humanoid robot Sophia pointed out, it’s not definitive that AGI would arrive by next year. In fact, some computer scientists deny the idea of AGI altogether. The reason is that there’s no clear definition of what AGI truly means. So, no one could say when exactly AGI will arrive because there’s no starting or finishing line to what these artificial intelligence systems could or should do.

However, this does shorten Goertzel’s timeline. Earlier, in 2019, he estimated that advanced machine intelligence would arrive in the next five to seven years, roughly between 2024 and 2026. But it has clearly not arrived yet. Thus, in an interview with Forbes, he revised the arrival to be around 2029 or 2030. But the recent advancements have made him positive about its arrival sooner than he or many others think.

Signs That AGI Is Not Here Yet

While Ben Goertzel is excited to see AGI arrive as soon as next year, he also acknowledged areas where AI systems still fall short.

One of the biggest causes of concern is the lack of long-term memory. There are many conversational AI solutions available today that can retain memory within a chat thread. Some are advanced enough to remember a few previous chats. However, no current AI system can potentially remember every chat with every user. Goertzel says that these systems should have a continuous personal history of their own. And they can do so when they are capable of updating their central parameters based on new experiences.

“They don’t have a long-term memory of their whole life like a person does. The result of that is they don’t know who and what they are and how they relate to the world around them.”

Ben Goertzel told IBM Think.

Even when such systems arrive, it will be challenging to prove if they are part of AGI. For that, Goertzel has proposed sending an AI robot to a college, which he calls the “MIT test.” If the robot can navigate the campus, study like a human student, interact with professors, and perform everything that humans do, it would likely be an advanced AGI system.

Another key concern Goertzel outlined earlier was ownership.

Who Owns Artificial General Intelligence?

Many LLM-based generative AI and agentic AI models are owned by businesses, government, or other organizations. However, ownership of AGI wouldn’t be that simple. A true AGI solution would behave more like a global infrastructure or an autonomous actor whose actions would cut through companies and even borders. Therefore, the question of ownership is both complicated and severe.

“I don’t want to give up control of what I’m doing to venture capital firms. Because I think AGI is too important for that.”

Ben Goertzel said on the On The Margin podcast, as reported in his Forbes interview.

Instead, he proposed a decentralized AI infrastructure. Through SingularityNET and the ASI Alliance, he advocates a blockchain‑based infrastructure for AGI. He aims to achieve something like what Linux and the internet have already proven. They are both open-source and publicly owned.

His company is already working on a blockchain solution for the same. He said that their “layer one ASI [Artificial Superintelligence] chain” should be ready for mainnet around the end of 2026.

Thus, businesses that want to capitalize on this will need to prepare their infrastructure, too. In fact, any AI system deployment is largely dependent on AI readiness to scale efficiently.

Ensuring public ownership is essential because governance is already the biggest issue with advanced artificial intelligence systems. Democratic legitimacy requires transparency, and private ownership cannot provide it, regardless of what it promises. Take the case of AI’s environmental impact. Many tech companies show the resources they use to run data centers to facilitate artificial intelligence tools and responses. However, many reports claim that the numbers are not accurate. Governance is way more sophisticated and ethics-oriented than resource consumption.

RAND, therefore, warns that if a single organization or government controls AGI, it will gain hegemonic power. It could simply crowd out competition and concentrate wealth and strategic leverage in the hands of a few.

“There is a tendency in our planning to confuse the unfamiliar with the improbable. The contingency that we have not considered seriously looks strange; what looks strange is thought improbable; what is improbable need not be considered seriously.”

Says economist Thomas C. Schelling, as covered by RAND’s How Artificial General Intelligence Could Affect the Rise and Fall of Nations report.

So, if you are wondering when AGI will arrive or how it would look, there’s no single answer to that. But while Artificial General Intelligence could still be years away, recent advances in AI agents and autonomous systems have made its arrival feel closer than ever. Similarly, the demand for AI chips to cope with the growing need for advanced AI tasks hints that we are moving in the right direction.

Ben Goertzel’s revised timeline reflects this growing optimism. However, it all depends on infrastructural readiness, long-term memory, governance, transparency, and ownership. Without infrastructural development, there will be no deployment and scaling. And if there’s a lack of governance, transparency, and ownership, there will be a trust gap, which would influence adoption negatively.

Frequently Asked Questions