Lim Woong: What we see when AI stares back

SEOUL, 13 OCTOBER 2026 —

Lim Woong, a professor at the Graduate School of Education at Yonsei University in Seoul, argued in a Korea Herald column published on October 12 that artificial intelligence should be understood as a mirror reflecting human nature rather than an autonomous force, and that accountability for the technology's harms must rest with the people who build, fund and deploy it.

Lim wrote that societies reach for metaphors — tireless assistant, parasite, oracle — to grasp AI's essence, but the metaphor a society embraces quietly shapes what it invests in, what it fears and what it chooses to regulate.

He proposed the mirror metaphor: the algorithms behind these systems are designed to reproduce aspects of human thought, while the vast quantities of human writing and behavioral data on which they are trained carry the imprint of our reasoning, our prejudices and our desires.

He acknowledged scholarly warnings that AI could threaten human safety and existence but said their roots lie not only in the machine's capabilities but in human appetites: developers racing for first-mover advantage, executives pursuing profit, and users eager to adopt the next new thing. Three examples illustrated how quickly AI's supposed neutrality collapses when human interests intervene.

In July, a Chinese robotics company launched what it billed as the world's first humanoid fighting league. In one bout, a robot reportedly continued swinging after its head came off. Lim wrote that the machine bears no grudge; it follows its programming and is expected to win.

The violence on display is therefore ours, staged and sold for entertainment, yet coverage can celebrate the spectacle and, almost in the same breath, voice concern about safety as though the entertainment and the danger belonged to separate worlds.

A Korean bank recently disclosed that customers' personal information, including names, addresses, phone numbers and other identifying details, was exposed in an incident attributed to malicious AI-assisted hacking. Subsequent reports suggested the impact extended to seven institutions and tens of thousands of people.

Lim noted that headlines marvel at AI's capacity to hack, turning a security breach into another story about the threat posed by AI itself, while far less attention goes to the people behind the attack, their motives or the defenses that should have been in place against a hardly unforeseeable threat.

Had a criminal gang stolen those records, the discussion would center on suspects, prosecutions and prison sentences; when an algorithm becomes part of the operation, it may change the speed, scale or cost of the crime but does not erase the human agency behind it.

Forged official documents, manipulated photographs and fabricated videos are established forms of deception that long predate generative AI, Lim wrote. People have always lied, distorted the truth and blurred the boundary between fact and fiction. Deception is one of our oldest habits, and the industries built around exploiting it are almost as old. The machine did not invent malice; it learned from ours.

Lim invoked a Korean storytelling motif: a man walks home at night, sees leaves trembling in the darkness and imagines ghosts; the next day, in a cheerful mood, he watches the sea toss those same leaves about and finds them beautiful. The scene changes little; the attitude changes everything. We began personifying AI almost as soon as it arrived, and now some would have us regard it as a god.

Much of our anxiety, he argued, comes from a limited understanding of our own capacity for harm and from our reluctance to confront what history has already taught us about ourselves. We have already committed atrocities on a scale that no AI has ever approached.

The central question, Lim concluded, is not simply whether AI will behave badly but who is building it, whose money is financing it, and what those investors and builders intend to do with it. A society that scrutinizes the machine while waving through its owners has put its magnifying glass over the wrong half of the picture. The reflection will improve only when its source does. The views expressed are the writer's own.

Related: Yonsei University · Lim Woong · Seoul