Afraid of the deal with the devil?

Choi Hoon

The author is the chief editor of the JoongAng Ilbo.

Humanity is turning into Doctor Faustus — a fictional scholar in a 16th-century play by Christopher Marlowe — who makes a deal with the devil to offer his soul in return for God-like unlimited power and knowledge, wrote Financial Times columnist Martin Wolf on his recent column on artificial intelligence. He defined humanity as “a species of competing with Doctor Faustuses, who seek knowledge and power, as he did,” adding, “But we cannot halt the bargains we have made […] We are [also] Mephistopheles [the demon in the legend of Faust]. The AI revolution will roll on.”

ChatGPT has caused a splash and shockwave across the world and industry. The advance of AI technology to the extent of mimicking and understanding human language as well as its nuances to provide detailed answers and perform human-like tasks, stirred excitement about the opportunities it can create as well as anxieties and fears about its replacement of human jobs and ethical challenges on top of the danger of spread in false and fake information and news.

In the first-ever U.S. Senate hearing devoted to AI technology, Sam Altman, founder and CEO of OpenAI, the creator of ChatGPT, praised the new generation of chatbots for their potential to “improve nearly every aspect of our lives” but demanded safeguards to “mitigate the risks of increasingly powerful models.”

Bill Gates, who helped the launch of the era of personal computers in the 1980s, thinks that AI chatbots like ChatGPT would have as disruptive impact on humanity as PCs and the internet. “In my lifetime, I’ve seen two demonstrations of technology that struck me as revolutionary,” he wrote in a blog post titled “The Age of AI has begun.” The first time was the graphic user interface demo in 1980, inspiring the future direction of Microsoft, and the second surprise was the first ChatGPT in 2022 when Gates challenged it to pass the Advanced Placement biology exam, which it passed within a few months.

Jensen Huang, the CEO of GPU giant Nvidia, called the ChatGPT the “iPhone moment” for AI, raising excitement about the democratization of computing and enabling everyone to become a programmer.

AI technology is defined by the AI parameters comparable to the number of synapses in the human brain. The GPT-3 released in 2020 has been trained with 175 billion parameters. The upgraded GPT-4 is presumed to have been trained with 1 trillion, or approximately 1 percent of 100 trillion to 150 trillion synapses in the adult human brain. “We cannot imagine what can happen when the AI technology arrives at five trillion to 10 trillion parameters or five to 10 percent of human brain capacity, and how fast that can happen,” said Kim Byoung-pil, a professor at KAIST. It could be the Maginot Line when the technology becomes beyond control to cause serious harm to humanity, as Altman warned.

The biggest concern is the impact on people’s jobs. According to “New Frontiers: The Origins and Content of New Work, 1940-2018,” co-written by MIT professor David Autor, 60 percent of global jobs in 2018 had not existed in 1940. Going back to Wolf’s article, he pointed out that the British had 3.3 million horses for pulling, transport, and cavalry. (In Korea, there were 40,000). But the mammal population is down by 75 percent today. “Could humans, too, become an outmoded technology, displaced by machines that are not just stronger and more dexterous but more intelligent, even more creative?” Wolf wondered.

Goldman Sachs report in April estimated that 300 million regular jobs could become vulnerable by AI. Two-thirds of American jobs could be automated to some degree, of which a quarter or up to half of the workload could be replaced by machines. On the upside, AI could raise labor productivity by 1.5 percent annually and global GDP by 7 percent ($7 trillion) over the next decade. Because most of the jobs are partially exposed to automation, their work could be more “compensated” rather than replaced.

In a recent article in Foreign Affairs, former Google CEO Eric Schmidt called AI as the game-changer of our time. “In previous eras, the technologies that shaped geopolitics — from bronze to steel, steam power to nuclear fission — were largely singular. There was a clear threshold of technological mastery, and once a country reached it, the playing field was leveled. AI, by contrast, is generative in nature. By presenting a platform for continuous scientific and technological innovation, it can lead to yet more innovation. Faster airplanes did not help build faster airplanes, but faster computers will help build faster computers. Technological advances in the next five to ten years will determine which country gains the upper hand in this world-shaping competition. The challenge for the United States, however, is that government officials are incentivized to avoid risk and focus on the short term, leaving the country to chronically underinvest in the technologies of the future.”

Korea ranks seventh in AI supremacy, trailing the United States, China, Britain, Canada, Israel, and Singapore. It is third in technology development and capabilities and sixth in infrastructure. But in talent, it is 29th and in an operational environment, 32nd. The country needs to upgrade its education system and engage aggressive immigration policy to draw talent. Overhauling in data regulations is also imperative. Korea must apply its specialty to smartphones by enabling faster AI applications on lighter and energy-efficient smartphones for practical AI services. If bargaining with the devil cannot be avoided, we must move faster.