Artificial Intelligence (AI) has come a long way since the concept was first introduced in the 1950s. Today, AI systems are ubiquitous, from personal digital assistants and chatbots to self-driving cars and predictive analytics. However, the field of AI is still in its infancy, and experts are divided on when we can expect to see truly transformative AI, such as Artificial General Intelligence (AGI) or Highly-Limited Machine Intelligence (HLMI). In 2011, computer scientist and mathematician Steve Omohundro made a prediction about the timeline for AGI, HLMI, and transformative AI. In this article, we’ll explore Omohundro’s prediction, his opinion on the intelligence explosion, and his vision for the future of humanity with AGI.
About Steve Omohundro
Steve Omohundro is a computer scientist and mathematician who has made significant contributions to the fields of artificial intelligence and robotics. He is the founder of the Artificial General Intelligence Research Institute (AGIRI) and currently serves as its president. Omohundro’s work on AI has focused on developing advanced algorithms for machine learning and intelligent decision-making. He has also explored the ethical and societal implications of advanced AI systems, particularly with respect to issues of safety and control. Omohundro has published numerous articles and papers on topics such as machine learning, robotics, and the ethics of AI. He has also served on several advisory boards and working groups related to AI, including the Future of Life Institute’s AI Safety Group and the Machine Intelligence Research Institute.
When asked about the timeline for AGI, HLMI, and transformative AI in a 2011 interview with Next Big Future, Omohundro replied, “n/a.” In other words, he did not make a specific prediction about when we can expect to see these types of AI. However, he did suggest that the development of AGI and transformative AI could potentially happen faster than some experts believe. According to Omohundro,
“the conditions for an intelligence explosion could be developing much faster than expected.”
Omohundro also noted that the development of HLMI systems could be easier and faster than the development of AGI. He suggested that HLMI systems could be developed by combining existing narrow AI systems, such as image recognition and natural language processing, with advanced decision-making algorithms.
Opinion on the Intelligence Explosion
The concept of the intelligence explosion is the idea that as AI becomes more advanced, it will become capable of designing even more intelligent systems, leading to an exponential increase in intelligence. When asked about his opinion on the intelligence explosion, Omohundro said, “I believe either scenario is technologically possible. But I think slower development would be preferable.” Omohundro noted that a sudden and uncontrolled intelligence explosion could lead to a catastrophic outcome. He suggested that careful planning and regulation would be necessary to ensure that advanced AI systems are developed with safety and ethical concerns in mind.
The Future of Humanity with AGI
Omohundro believes that rapidly bringing about AGI is not the best initial goal. Instead, he believes that a clear roadmap for safely integrating these systems into society is needed. In his own words,
“I would feel much better about it if we had a clear roadmap for how these systems will be safely integrated into society for the benefit of humanity. […] I believe the best approach will be to develop provably limited systems and to use those in designing more powerful ones that will have a beneficial impact.”
Omohundro suggested that provably limited systems, which are AI systems that are designed to operate within specific boundaries and limitations, could be used to develop more advanced AI systems. He believes that this approach could help ensure that AI systems are developed with safety and ethical concerns in mind.
The Future of Humanity with AGI/HLMI/Transformative AI
Given the potential power and complexity of AGI, it’s understandable that many people are concerned about the risks of developing such technology. Omohundro has been vocal in his opinion that rapidly bringing about AGI is not the best initial goal. Instead, he believes that the development of provably limited systems that can be safely integrated into society is a better approach. As Omohundro notes, designing AGI with safety and control in mind is critical to ensuring that the technology benefits humanity rather than causing harm. A clear roadmap for safely integrating these systems into society is essential, and it’s important that ethical considerations be taken into account at every stage of development. One potential approach to developing safe and beneficial AGI is through the use of “provably beneficial” AI systems. These systems are designed with an objective function that explicitly incorporates human values and goals, which can help ensure that the AI acts in ways that are aligned with human interests. Another potential approach is to develop “value-aligned” AI, which is designed to pursue goals that are consistent with human values and preferences. This can be achieved through the use of machine learning algorithms that are trained on large datasets of human preferences and behaviors. Regardless of the specific approach taken, it’s clear that the development of safe and beneficial AGI is a complex and challenging task. However, as Omohundro notes, it’s a task that is well worth pursuing given the potential benefits that AGI could bring to society.
In conclusion, Steve Omohundro’s prediction that AGI could be achieved within the next few decades has continued to influence discussions about the future of AI and its potential impact on society. While the development of AGI is a complex and challenging task, Omohundro’s work has highlighted the importance of developing safe and beneficial AI systems that are aligned with human values and goals. As AI continues to advance and become more integrated into our daily lives, it’s important that we continue to prioritize ethical considerations and ensure that AI is used to benefit humanity. https://youtu.be/Kb2HzU5Ti30