Chat GPT's answer "bureaucrats will eventually scrutinize"

 

Chat GPT's answer "bureaucrats will eventually scrutinize", fear of entering undisclosed information ... Voices of doubt about the introduction of answers

                                                                                    Chat GPT input screen

Minister of Economy, Trade and Industry Nishimura announced at a press conference after the cabinet meeting on the 11th that interactive AI (artificial intelligence) "Chat GPT Regarding the possibility of preparing an answer to the Diet, some experts have voiced concerns about the security of information management and pointed out that the effectiveness of the introduction is questionable.

"In the end, bureaucrats will have to scrutinize the chat GPT responses." Mr. Nishimura lists the reason for consideration as "Civil Servants Former bureaucrats Takao Komine - A visiting professor at Taisho University showed such a view.

Komine said that while there are many advantages to using chat GPT for bureaucratic work, such as researching systems in other countries, he also pointed out, "There are cases where answers can be forcibly created from public information, and it is dangerous to just make things up.". "It is not possible to entrust the whole of the government's official statements to the Diet to the ChatGPT," he said.

Ichiro Sato, a professor of informatics at the National Institute of Informatics, said, "There is a risk that bureaucrats will enter non-disclosed information when preparing answers, and that it will be used for learning chat GPT, and confidential information will be leaked." points out.

Chat GPT reads a large amount of online information and generates new sentences in response to instructions. Even if the content is incorrect, it is difficult to notice because of the natural sentences, and there is a risk that false information will spread.

In addition, in the learning data copyright, if a certain sentence is included, there is a possibility that the expression is the same as or similar to that sentence and infringes copyright. Depending on the learning data, there is also the risk of returning answers that encourage discrimination and prejudice.

There are also voices of concern about the negative impact of the educational world. Chat GPT is the thesis and report. In a congratulatory address at the graduation ceremony on the 27th of last month, Nao Sugiyama, president of Nagoya University, said, "The situation can be said to be a crisis for university education."

A series of universities, such as the University of Tokyo, has announced policies to restrict the use of student reports. Sophia University has notified students that they will not allow the use of chat GPT for purposes such as writing reports without permission. A person in charge at the university said in an interview, ``From the perspective of ensuring the quality of education and fairness, creating with generative AI is not acceptable in the same way as stealing or asking others to create.''

The Ministry of Education, Culture, Sports, Science, and Technology also plans to create materials that show how it is handled in educational settings.


No comments

Powered by Blogger.