AI 가짜뉴스 초고

[김정후]

Yongsan can also burn at any time. Last year, one photo became a big controversy in the U.S. The photo captured the scene of a big fire in the Pentagon. The image spread quickly through social media, and plunged the U.S. stock market into great confusion. Gold and government bond prices also went up significantly. However, the photo was a "fake" image created by artificial intelligence. Lies were produced indiscriminately and circulated indiscriminately. The problem is that influential media, such as Bloomberg, also reported the photo as a fact. In short, if the media crisis in the multimedia era was a matter of "influence," the media's “credibility” has recently declined.

The media crisis that emerged with AI took a very different aspect from the discussions so far. This is essentially a different matter from the lowest level of trust in Korean media in OECD countries. Until now, the media has lost trust by advocating excessive factionalism, but it has fallen into the category that can be improved with the human will. In other words, media gatekeeping was structurally effective. On the other hand, recently, the media has been threatened with the reliability of artificial intelligence. This is because the fake facts created by AI based on big data, not 'selective' facts of humans, are outside the control of the media.

Deeplearning collects geometrically large amounts of information. That's why it's impossible to pinpoint the source of the information. Media gatekeeping is being disabled due to technological advances that go beyond the facts of information. Consider, for example, a photo of former President Trump and Vatican Pope, who were recently arrested by police, wearing expensive outerwear. At first glance, it's a plausible lie that seems too realistic. Nevertheless, the media reported the photo as true. What's more problematic is that the history of the technological revolution so far proves that humanity will not return to the pre-development state.

In the era of rapid technological development, the way the media recovers its credibility inevitably meets the question of how to use technology. In a nutshell, there is XAI. It is a technology that is actively used in the actual industrial field and a new form of artificial intelligence that can clearly identify the source of information. Why don't you use it for media? In a nutshell, if the media can manage the source of information generated by the AI, it will be possible to prevent the spread of fake data indiscriminately. Just as the media in the multimedia era has created the concept of news 'content', an alternative using technology in the generative AI era is needed.

[UNGP]

Lies were produced indiscriminately and circulated indiscriminately. > 이 부분에서 겹치는 부사는 한 번만 써주면 어떨까 합니다.
전반적으로 디테일한 사례들이 많아 좋았습니다. 언론이 어떤 역할을 해야할지 대안제시가 조금 더 구체적으로 되었다면 어떨까 합니다. 잘 읽었습니다!

[우지원]

서론의 첫 문장이 인상 깊었습니다. 첨언하자면 언론의 역할이 조금 더 자세하게 작성되면 좋을 거 같은데, 마지막 문단에서언급하신 XAL에 대해 더 디테일하게 설명해줘도 좋을 거 같습니다. 잘 읽었습니다. 고생하셨습니디:)

[송정현]

전체적으로 좋은 글이었습니다. 가짜뉴스의 예시와 위험성을 잘 서술해주었으나 언론사의 역할을 논하는 부분이 맨 끝에 등장해 조금 아쉬었습니다. 퇴고에서는 언론사의 역할을 핵심적으로 다루어야 할 것 같습니다. 배경지식을 조금 줄이고 앞으로의 언론사의 역할과 책임성을 조금 더 조명해주신다면 더 균형있는 에세이가 될 것 같습니다. 언론사 또한 AI를 뉴스에 활용하자는 제안이 신선했습니다.

그리고 minor한 부분이지만 코리아헤럴드에서는 명칭을 먼저 풀어쓰고 그 다음에 약자를 씁니다. 예를 들어 두번째 문단에서 "OECD" 를 처음 언급할 때는 먼저 다 풀어서 써주시고: OECD> Organisation for Economic Co-operation and Development 그 이후로는 OECD 약자를 쓰시면 됩니다. 해당 언론사들의 명칭 표기방식 같은 걸 숙지하시면 나중에 플러스 요인이 되지 않을 까 생각합니다.