[경제]’Gender discrimination controversy’ AI chatbot’Achieved’ service temporarily suspended

[경제]’Gender discrimination controversy’ AI chatbot’Achieved’ service temporarily suspended

by

‘Deep learning’ based on conversations with real lovers, etc.
Service discontinued after controversy over gender discrimination and hate
‘Achieved’ developer suspects user personal information leakage

[앵커]

The AI ​​chatbot’Iruda’, which was caught in controversy over hate speech and personal information leakage, eventually temporarily stopped the service.

The developer apologized for the various controversies and said it would have a period for improvement.

This is Yeom Hye-won.

[기자]

‘Iruda’, an AI chatbot that explicitly expressed negative intentions, saying that pregnant women on the subway were disgusting, and exchanged homophobic remarks and even sexual conversations.

‘Iruda’, applied with the’deep learning’ technology in which computers think and learn like humans, learned by themselves based on conversations between real lovers.

Based on this, it became popular with more than 400,000 users within 3 weeks of launch, but after controversy such as a series of hate speeches, the developer eventually temporarily stopped the service.

Developer ScatterLab said it sincerely apologizes for any cases of discriminatory remarks against certain minorities.

However, another controversy about the leakage of personal information was drawn.

In response to the problem that the company’s other service users’ personal information was used without sufficient explanation for the development of’Achieved’, the company apologized for not sufficiently communicating with users, but said that information that could identify individuals has been removed. I explained.

With the controversy over’achieved’, voices are gaining momentum that discussions about the ethics of artificial intelligence development are necessary.

As the subjectivity of artificial intelligence designers and the ethical consciousness of users can all be a problem, the government also announced an artificial intelligence roadmap at the end of last year to strengthen education.

[강도현 / 과학기술정보통신부 인공지능기반정책관 (지난해 12월) : 인공지능 윤리 기준을 마련함에 따라 이후 후속적으로 윤리교육의 커리큘럼을 연구 개발하고 각계각층에 맞는 커리큘럼의 개발과 보급들을 진행할 계획으로 있습니다.]

Due to the nature of self-learning artificial intelligence, simply increasing the number of taboos cannot be a fundamental prescription.

It is pointed out that concrete discussions on how to develop and use artificial intelligence in accordance with the international AI ethical standards that use high-quality data to develop intelligence that is beneficial to humans are needed.

This is YTN Yeom Hye-won.

[저작권자(c) YTN & YTN plus 무단전재 및 재배포 금지]

.Source