In this blog post, we’ll explore the possibilities of the superintelligence era and the steps humanity needs to take to prepare.
Through three industrial revolutions to date, humanity has created a world entirely different from anything that came before. And now, we are on the cusp of yet another industrial revolution. The core technology driving this revolution is artificial intelligence (AI). AI has established itself as one of the most prominent technologies in recent years, and public interest has surged significantly, particularly since 2016, when AlphaGo defeated Go master Lee Sedol.
So, to what extent has artificial intelligence advanced today? AI is broadly categorized into weak AI and strong AI. The most significant difference between the two lies in the presence or absence of self-awareness and general-purpose thinking capabilities. Weak AI refers to systems designed to solve problems in specific domains without self-awareness; it is used to compensate for human limitations and enhance productivity. In contrast, Strong AI refers to systems with a sense of self that can think freely and solve a wide range of problems like humans; it is generally also known as Artificial General Intelligence (AGI). All AI systems put into practical use to date fall under the category of Weak AI and remain limited to demonstrating specialized performance in specific fields.
Weak AI is already being utilized in various fields. Google’s self-driving technology, centered around Waymo, has advanced to the point where commercial services are operating on public roads, and in the medical field, AI is being used in diverse areas such as medical image analysis, diagnostic assistance, and new drug development. Furthermore, voice assistants—including Apple’s Siri—and generative AI services are already being used naturally in everyday life.
Humanity has already entered an era of coexistence with AI. Moreover, the pace of AI development continues to accelerate. Just as computer technology has advanced by leaps and bounds over the past few decades, AI is also projected to develop even more rapidly in the future. Some researchers predict that once AI reaches a stage where it can improve its own performance, the rate of its intellectual development could increase exponentially. Based on this possibility, some scholars are discussing the emergence of “superintelligence”—intelligence that surpasses human intelligence.
Many scholars and experts are offering various perspectives on the future direction of AI development. Futurist Ray Kurzweil has predicted the emergence of general artificial intelligence and superintelligence, raising the possibility of a “technological singularity.” On the other hand, Nick Bostrom has mentioned the possibility of superintelligence surpassing humans and has consistently emphasized the need for safety and control measures to prepare for it. However, these forecasts are predictions about a future that has not yet materialized, and to date, neither general artificial intelligence nor superintelligence has become a reality.
Many experts predict that the emergence of superintelligence will fundamentally transform human life. Bill Gates, Elon Musk, and others have warned that developing superintelligence without sufficient safety measures could pose a serious threat to humanity. They express concern that if an AI with intelligence superior to humans were to escape human control, dystopian scenarios like those depicted in the movies ‘Terminator’ or ‘The Matrix’ could become a reality.
On the other hand, Kurzweil offers a slightly different perspective. He predicts that as AI advances, humans will integrate it with their own bodies to gain even greater capabilities. For example, he suggests the possibility that the convergence of nanotechnology and AI could enhance human cognitive abilities and significantly improve health and longevity. Of course, these predictions also pertain to a future that has not yet been realized.
Will a future with superintelligence be a utopia or a dystopia? No one can say for certain. However, to create a more positive future, efforts to develop AI safely and responsibly are absolutely essential. Bostrom emphasizes that institutions and norms capable of appropriately regulating AI must be established. Ultimately, developing AI based on ethics and safety is arguably the most important way to mitigate the risks posed by superintelligence.
The era of superintelligence may one day become a reality for us. It could become an entity that enriches human life, like Samantha in the movie ‘Her’, or it could become a threat to humanity, like Ultron in ‘Avengers: Age of Ultron’. It is still unclear which future will become reality. What is certain, however, is that the ethical awareness of AI researchers and the sustained attention of society as a whole will be crucial factors in determining the path forward. Humanity must not only advance technology but also develop the ethics and systems necessary to use that technology responsibly.
