The individual who introduced the term ‘artificial general intelligence’ (AGI) and is the CEO of a company, thinks that AI is on the brink of experiencing a rapid ‘intelligence explosion.’
Ben Goertzel, a PhD mathematician and futurist, recently predicted at a summit on AGI that we could potentially achieve human-level AGI within the next three to eight years.
According to Goertzel, often referred to as the ‘father of AGI,’ once human-level AGI is achieved, a significantly more advanced AGI could be developed within a short period.
The futurist acknowledged the possibility of error but confidently forecasted a scenario where an extremely advanced AI surpasses its human creators, with the only potential obstacle being the AI’s own cautious nature.
Last week, at the ‘2024 Beneficial AI Summit and Unconference,’ Goertzel gave his forecasts in a speech that was partially supported by SingularityNET, the company of which he is the CEO.
During his speech at the event, which was hosted this year in Panama City, Goertzel admitted that there are known unknowns and likely unknown unknowns.
“Nobody knows for sure when we’re going to reach human-level artificial general intelligence [AGI]; nobody has created it yet.”
However, Goertzel believed that until a “quantum computer with a million qubits or something” was needed for the processing power, an exponential increase in AI was unavoidable.
He stated, “In my opinion, once you reach human-level AGI, you could get a radically superhuman AGI within a few years.”
Goertzel has been researching “artificial super intelligence” (ASI) in recent years. ASI, according to Goertzel, is an artificial intelligence (AI) that is so sophisticated as to rival the combined brain and processing power of human civilization.
Goertzel stated that his argument is supported by “three lines of converging evidence.”
First, he referenced the most recent research by Ray Kurzweil, Google’s longtime resident futurist and computer scientist, who created a prediction model indicating that AGI will be possible by 2029.
Kurzweil’s theory, which he will expand upon in his upcoming book “The Singularity is Nearer,” was informed in part by statistics demonstrating the exponential nature of technological advancement in other tech-related industries.
Goertzel then listed all the well-known recent advancements in “large language models” (LLMs) during the last several years, claiming that these developments had “woken up so much of the world to the potential of AI.”
Eric Schmidt, the former CEO of Google, is advocating for global cooperation among governments to establish regulations for AI to prevent it from becoming uncontrollable. At a summit last November, Schmidt drew a comparison between the importance of regulating AI and the need for weapon regulations following the bombings of Nagasaki and Hiroshima in WWII.
Finally, the computer scientist, wearing his distinctive leopard print hat, focused on his infrastructure research project that integrates different AI infrastructure components, known as ‘OpenCog Hyperon.’
The new infrastructure aims to integrate advanced AI technologies, such as LLMs, with new forms of AI focused on various cognitive reasoning areas like math, physics, and philosophy, to develop a comprehensive AGI.
Goertzel’s project, ‘OpenCog Hyperon,’ has received support and attention from individuals in the AI field, such as Berkeley Artificial Intelligence Research (BAIR), who recently published an article co-authored by Goertzel, Databricks CTO Matei Zaharia, and others.
This isn’t the first alarming or confident prediction on AI that Goertzel has made in recent years.
According to a futurist in May 2023, AI has the potential to take over 80 percent of jobs within the next few years.
He mentioned at the Web Summit in Rio de Janeiro that month that nearly all jobs involving paperwork could be automated.
Goertzel mentioned that he did not view this as a negative, stating that it would enable individuals to discover more fulfilling pursuits than simply working to make a living.
In that same month, he also mentioned to the site Futurism: ‘I have used substances with an AI, in the sense that I have used substances and then engaged with an AI.’
One of the eccentric episodes in Goertzel’s history is the ‘psychedelic’ practice, which was part of his work on ‘algorithmic music composition’ in the 1990s.
The individual, who describes himself as a panpsychist and claims that even inanimate objects possess consciousness, has proposed that researchers work towards developing a ‘benign superintelligence.’
Goertzel has suggested an AI-powered cryptocurrency rating agency that can detect fraudulent tokens and coins.
However, the computer scientist is widely recognized for his contributions to Sophia the Robot, the first robot to receive legal citizenship.