Artificial intelligence technologies that were only seen in existing movies
Expanding from research to everyday life

At CES 2016, IBM President Virginia Rometty predicted that AI (artificial intelligence), a technology that thinks and judges like humans, will lead the future of the IT industry. She explained, “We cannot create differentiated value in the huge Internet of Things market with just digital technology,” and “We need artificial intelligence technology that understands and infers data for data analysis that brings business insights.” Artificial intelligence has always been the center of attention, as seen in many movies. This is not a new story. But why is it being talked about so much this year?

Professor Park Byeong-jun of the Department of Computer Software Engineering at Kwangwoon University pointed out, “If you look at the history of artificial intelligence, it is one of the fields that has had severe ups and downs.” He explained, “It was mainly focused on research and development, but interest is growing as the research results are now developing to the point where they can be used in everyday life.”

The artificial intelligence 'JARVIS' from the movie 'IRON MAN'

When thinking about artificial intelligence, most people think of 'robots'. From the female robot 'Maria' in the 1927 film 'Metropolis' to the powerful robot 'Terminator' and Steven Spielberg's work 'AI Robot', movies that gave people both freshness and shock.

In recent SF movies, it is often the OS (Operating System) that appears, not the robot. This can be seen by looking at the popularity of 'Samantha' from the movie 'Her' and 'Javis' from Iron Man. This operating system is much smarter than humans and appears as an indispensable friend who can perceive not only information and knowledge, but also human emotions.

Expert systems, which are ‘weak artificial intelligence’, are already actively used

The history of artificial intelligence begins in 1955, when the term artificial intelligence was first used in a research plan by Marvin Minsky, a neuroscientist, and about ten others to teach computers human intellectual activities.

After that, already in the 1970s, expert systems were introduced that could analyze and judge based on advanced knowledge, such as 'MYCIN', which collected and applied the knowledge of infectious disease specialists, and 'POSPECTOR', which specialized in finding gold in mining development terrain, accurately and quickly processed tasks that took a long time. These artificial intelligences, which are still widely used today, are called 'weak artificial intelligence' and are technologies that we can commonly encounter in our daily lives.

At that time, many scientists thought that they could quickly create a system that could think and act like ordinary people by seeing how highly specialized tasks were being performed. However, it was not as easy as they thought. An expert system is one that inputs all the information necessary for a robot in advance through 'top-down injection'. Although it is possible to input information about the specific field mentioned above, it is impossible to input all the information in everyday life that is highly variable.

So we looked for another method of information input, and it was the 'deep learning' technology, a 'bottom-up command' method.

In fact, deep learning technology itself was not recently announced. As early as 1943, the artificial neural network algorithm similar to the human brain structure was first introduced by Warren McCulloch, an associate professor of psychiatry at the University of Illinois in the United States, and in the 1980s, the deep neural network theory with multiple hidden layers between the input layer and the output layer emerged. This deep neural network theory can be said to be the current deep learning. Just as humans become adults through education and experience over a long life, it was intended to grant the ability to learn and make judgments on their own in addition to commands that are not input through big data.

Artificial intelligence being applied to the medical and financial industries

Deep learning in the medical industry is one of the high-tech technologies that has recently been attracting attention even among startups. LG Research Institute explained that 'Deep Genomics' aims to identify mutant/unique genes through pattern analysis from a genetic perspective and suggest customized disease treatments accordingly, and said that it started with a vision of enabling effective new drug development and customized medicine.


The main character of the movie ' Transcendence ' becomes an artificial intelligence inside a computer.

Also, in the movie Transcendence, there is a scene where the main character's brain is uploaded to a computer and evolves into artificial intelligence, which then generates huge investment returns. This is not just a story in a movie. Artificial intelligence is already being applied in financial services.

According to LG Research, Development Bank of Singapore (DBS) recently announced that it is using IBM's Watson in its asset management business to provide customized investment advice and asset management services to its top clients. Additionally, the Australia and New Zealand Banking Group (ANZ) has signed a $450 million contract with IBM to improve the quality of its investment advisory services, and South Africa's Nedbank is looking to use Watson for areas such as social media monitoring.

The movie 'Prometheus', which describes the origin of mankind as an engineer

The movie 'Prometheus' is a movie that shows the desire to find the origin of humans and create 'something similar' to themselves. In the movie, the ancestors of humans are named 'engineers', which equates humans with scientists and technicians, or engineers, who want to create robots based on artificial intelligence. Why do we want to create artificial intelligence and robots that are like humans, not humans?

Artificial intelligence that is similar to humans is classified as ‘strong artificial intelligence’. In the play, David, an android robot, says, “I am programmed not to feel rejection from humans.” One of the issues with strong artificial intelligence is to prevent people from recognizing the artificial intelligence when talking to it.

One of the features that makes it less intimidating is the part about natural language usage. Professor Park Byung-joon said, “Even in the field of artificial intelligence, the slowest field is understanding and speaking natural language,” and expressed his opinion, saying, “When you listen to the stories of those who are actually researching the field, you become skeptical.”