[ad_1]
The portal side insisted that “people do not intervene in the editing and arrangement of the news”
Arguing around the world about the ‘neutrality of artificial intelligence’
Founder of Daum “I am wrong when I say that artificial intelligence is objective”
At a plenary session held at the National Assembly in Yeouido, Yeongdeungpo-gu, Seoul, on the 8th, when the Democratic Party representative, Yoon Young-chan, received a call for the speech by the negotiating group Joo Ho-young to be introduced to the main Kakao Talk via a chat room with his parliamentary office staff, ‘Kakao is too much. He’s writing “Go ahead.” Joint coverage photo
When the controversy arose when a controversy arose from Yoon Young-chan, a member of the Democratic Party, who pressured Kakao in relation to the main news of the portal site Daum, Kakao, which operates Daum, officially announced the possibility of editing saying, “Artificial intelligence (AI) is organizing news.” I came out with my denial. However, even if artificial intelligence organizes the news, doubts are constant as to whether there is no room for an “ artificial ” adjustment by the staff within the portal, and whether the artificial intelligence work fully ensures objectivity and neutrality remains controversial.
■ Naver and Kakao, “The news arrangement is in charge of artificial intelligence” Regarding the controversy over the “main news article exposure” that Senator Yoon lit on the 8th, Kakao said, “The news displayed on the main screen is automatically organized by artificial intelligence as of 2015.” The article is different. “This means that the main screen exposed to consumers who enjoy political news and BTS fans who enjoy entertainment news are different. Kakao offers news services on the Daum portal and the tab of KakaoTalk store. Among the 30,000 articles that the media sends to Kakao every day, artificial intelligence is responsible for inferring between 400 and 600 articles by filtering duplicate / advertisement articles and sensational articles. Kakao explained that the layout of news depends on the recommendation algorithm based on gender, age group and user interests. In a conversation with CEO Kakao, “It was already in 2015 when the system was built so that no human intervention was involved in editing and news organization. It is a structure that makes human intervention impossible. ”Naver also It has been reorganized so that the news does not appear on the main screen of the Naver application since October 2018. The news tab, which can be seen by swiping to the right of the main screen, is managed by displaying articles edited by the media or automatically recommending news of interest (MY News) by artificial intelligence. Naver’s side explained: “To avoid confirmation bias, as well as articles that people would prefer, the algorithm is designed to recommend articles that normally don’t look good.”
The founder of the Daum portal, Lee Jae-woong, former CEO of Soka, posted on his Facebook the night of 8 Click on the image to enlarge it.
■ Is artificial intelligence really value neutral? However, there is an opinion that in the industry there is still a way for humans to artificially intervene in the editing and arrangement of news in ‘special’ situations such as artificial intelligence errors. For example, artificial intelligence may leak news that should be exposed to everyone, such as natural disasters. In this regard, CEO Lee Sang-ki said, “If the machine caused an error and the items that should have been leaked are not leaking, we will repair the machine. “There is no case that humans can even correct the mistake,” he said. “It is said that it is impossible to systematically say that nobody is doing the job.” Apart from this, it is also controversial whether the provision of artificial intelligence news ensures objectivity and neutrality. On the night of the 8th, former Soka CEO Jae-woong Lee, the next founder, said: “Many people think that artificial intelligence is value neutral, but it is not. He pointed out that the portal that replied that ‘we are neutral because artificial intelligence did it’ is as irresponsible as Congressman Yoon’s protest. This means that the design of artificial intelligence algorithms is also a work of humans, so it will surely be influenced by human value judgment. Discussions about the neutrality of artificial intelligence have been around for a long time. In April of last year, the Software Policy Research Institute said: “There is a growing voice of concern about the negative impact of digital technologies such as artificial intelligence, big data and the Internet.” “The main reason people are concerned about using artificial intelligence is that artificial intelligence is fairer than expected. Because I don’t. “Professor Michael Zimmer of the University of Wisconsin-Milwaukee, an expert in artificial intelligence, also noted:” It is easy to confuse artificial intelligence with neutrality, but it will be a means of maintaining prejudice and injustice. forever existing society. “By Choi Min-young, Staff Reporter [email protected]
[ad_2]
