Using the example of the movie “Avengers: Age of Ultron”, we discuss the impact and possibilities of artificial intelligence on humanity. The rapid development of AI could be a critical factor in determining the future of humanity, emphasizing the need for international cooperation and philosophical exploration.
In the movie “Avengers: Age of Ultron” (released in 2015), the protagonist Tony Stark develops an artificial intelligence to protect Earth from aliens. However, the artificial intelligence “Ultron” decides that humanity must be destroyed to protect Earth and attacks humanity. The idea that A.I. could attack humanity isn’t just a figment of the movie’s imagination. It’s a topic that has been discussed by experts. For example, in December 2014, Stephen Hawking warned the BBC that AI could lead to humanity’s destruction. Of course, not all experts believe that artificial intelligence (AI) will lead to the end of the world, but it is worth considering the possibility and how to deal with it.
The concept of artificial intelligence first emerged in 1956 at the Dartmouth Summer Research Project, a workshop organized by computer scientist John McCarthy. Since then, engineers have been trying to create artificial intelligence with the same or better intelligence than humans. Currently, AI is still in its infancy compared to humans, but it is rapidly improving. The current state of artificial intelligence is already being utilized in various fields. For example, A.I. is used as an automatic translator in Naver, South Korea’s leading search engine, and it is also behind speech recognition technology. In early 2016, the entire country witnessed AlphaGo, a Go-playing artificial intelligence.
There are several limitations to our current understanding of A.I. First, each A.I. is currently used only in specialized areas, which means it is not diffusible. Once an A.I. has a role, it cannot do anything outside of that role. For example, AlphaGo is a Go AI, so it can only play Go, not act as a translator. Second, A.I. is dependent on humanity. All A.I. is subject to humans and has no agency.
However, it is only a matter of time before these limitations are overcome, not a matter of possibility. Just as smartphones have reached nearly every citizen in South Korea in the decade since the iPhone was introduced in 2007, AI is expected to advance rapidly. According to Nick Bostrom, director of the Institute for the Future of Humanity at Oxford University, by 2040-2050, AI will be 50 percent as intelligent as humans in all areas, and by 2075, it will be 90 percent. The neurons that make up the human brain can send information two hundred times per second, while modern computers can send gigabytes at a time. Furthermore, the speed at which neurons send information is a hundred meters per second, whereas computers send information at the speed of light. In other words, A.I. is mathematically infinitely more capable than humans.
With such unlimited possibilities, A.I. is free from humanity. In this case, there is also the prospect of humanity being bound by AI. Sam Harris, a leading American commentator and neuroscientist, emphasized that such a development is inevitable. This is when the “intelligence explosion” occurs. The intelligence explosion is the moment when A.I. exceeds human intelligence and creates a better species than humanity. In other words, A.I. that is smarter than humanity will develop further on its own and escape from human control. So the predictions of some experts that humanity is doomed are somewhat plausible. Others argue that an intelligence explosion is the final step toward utopia.
The reason behind the conflicting claims is that humans have different ideas about the purpose of AI. The pessimistic view of AI is that humanity will set its purpose in a way that harms humanity, while the optimistic view is that AI will set its purpose in a way that helps humanity thrive. In other words, how we set AI’s purpose will determine the future of humanity. The key is not to pursue short-sighted goals like the Hand of Midas, but to set the initial value of AI after the intelligence explosion as the value that humanity should universally pursue or the means to enable all human beings to prosper without discrimination.
What the initial value of AI should be is still an open question. However, it should be an ongoing discussion. This is because if the AI explosion and this problem are not solved at the same time, there will be irreversible consequences. Currently, our society thinks that this is a movie, and we don’t have a serious discussion about it.
Even before AI, we developed a technology that could end human history. Nuclear weapons. At the time of their creation, physicist Albert Einstein predicted that nuclear weapons would be the last invention of mankind. He warned of their dangers, saying that they would make it impossible for humans to survive on Earth. But despite the invention of such dangerous weapons, we are thriving on Earth. Nuclear weapons have not been used since the 1950s. In addition, the technology used to make nuclear weapons has been used to create nuclear power plants, an alternative energy source to oil. Similarly, to prepare for future intelligent detonations, we can learn from past precedents on how to control and manage nuclear weapons. Nuclear weapons are internationally recognized, meaning that nearly all nations have cooperated, debated, and worked together to find a way forward to save humanity from the dangers of nuclear weapons. As with nuclear weapons, preparing for an IED requires collaboration through the formation of international organizations.
In addition, to prepare for a post-intelligence explosion, it is necessary to consider what are the universal values of humanity. The development of artificial intelligence requires not only engineering calculations but also humanities, which means philosophical exploration of human beings and thinking about the reason for human existence. It is only when engineers have such concerns that they need a period of time or education that allows them to develop not only engineering knowledge but also humanities. For future engineers, universities need to teach humanities courses to engineering students, and engineering students need to develop a humanistic attitude.
Artificial intelligence that is more intelligent than humans will be the last invention of mankind. If the initial value setting is correct, it will open a utopia by doing the dirty work of humanity after the intelligence explosion, and it will also create inventions that are beyond the scope of human imagination. If the default settings are off, the target of the AI’s attack will be humanity. In this case, a more intelligent AI could wipe out humanity. It’s up to us to decide whether the last invention will end our future or allow it to flourish.
