[Namsan Ddalggakbari] Benevolent Artificial Intelligence? The Answer Lies in Human Ethics

Introduction to AI Ethics / Communication Books / Ko Youngsang et al. / 18,810 KRW

[Asia Economy Reporter Kim Hyojin] Amazon, the world's largest e-commerce company, completely discarded its data-driven artificial intelligence (AI) recruitment system after it caused errors by recommending only men for most developer and technical positions. The AI-based crime prediction program COMPAS faced criticism for its bias, predicting the recidivism rate of Black offenders to be more than twice that of white offenders.

Ha Yoon-soo, former president of Busan National University of Education and chairman of the Korean Federation of Teachers' Associations, recently pointed out at the 'AI KOREA 2021 Artificial Intelligence Ethics Expo' conference hosted by Busan City that "advanced AI and metaverse technologies are gaining attention as new technologies that will lead future society and educational revolutions, but at the same time, misuse and crimes such as AI bias and deepfakes are occurring frequently."

Chairman Ha also stated, "Not long ago, Microsoft's AI chatbot Tay spewed hateful remarks including racism and sexism, leading to the suspension of its service. There have also been cases where deepfake technology was abused during remote classes, violating teachers' human rights," urging the prompt establishment of ethical standards related to AI use for valuable coexistence between humans and AI. He believes that concerns about dystopia are growing as much as the utopian expectations AI may bring.

The 'gateway' to the AI era, which was once thought to be 'still far off,' is now rapidly opening in conjunction with the unprecedented situation of COVID-19. The days when AI Go genius AlphaGo defeated human Go masters are already in the past. AI kindergarten teachers are educating children on alphabets and numbers, teaching songs, drawing, yoga, and gymnastics, while AI doctors and AI companion animals are actively taking care of human health and emotions.

There are also forecasts that AI idols will soon shake up the global popular culture landscape. Regardless of their form, in the not-too-distant future, every household may enter an era of 'one robot per person.' The problem is that AI is nothing more than 'intelligence.'

Reference Image

Including the aforementioned issues of data bias and misuse as criminal tools, problems such as invasion of users' privacy during the learning process and the development of so-called 'killer robots' that could pose fatal threats to humans can be raised endlessly. This means that the challenges AI presents to us are as broad and deep as bridging the gap between AI and 'humanity' and even protecting human life.

AI is merely a means for human happiness and benefit, but can the imperative that humans are the sole and ultimate 'purpose' truly be upheld? Can humans and humanity as 'purposes' be protected from the logic of economic prioritization arising from the advancement of industry and capital markets? The book Introduction to Artificial Intelligence Ethics was published by 11 experts from the Korean Artificial Intelligence Ethics Association with the intention of sharing these concerns and seeking solutions.

"Creating benevolent artificial intelligence is not a simple matter. It is not resolved by simply inputting the command 'make humans happy' into AI. In fact, such a simple command could lead AI to blindly conclude, 'After learning all the online data existing on Earth, the beings that make humans most unhappy are humans themselves. Therefore, to make humans happy, humans must be exterminated' (from the text)."

'Artificial intelligence ethics' refers to the ethics humans must observe in the AI domain. If AI advances too quickly, side effects will follow accordingly, but if AI development is excessively suppressed, technological progress beneficial to humans may be hindered. Therefore, what is important is the harmonious development of AI advancement and AI ethics. The point where we stand now may be the starting point of such a grand discourse. Besides the profound considerations about AI, Introduction to Artificial Intelligence Ethics also incorporates various interesting AI stories that we may not be well aware of.

It also touches on speculative science fiction questions such as how to view violence humans inflict on humanoid robots, whether a robotic arm equipped with AI technology can be considered part of the human body, whether damaging a robotic arm should be treated as injury or theft, and if a human body becomes a cyborg, whether the person treating that body is a doctor or an engineer.

This content was produced with the assistance of AI translation services.