Chinese chatbot service ‘Chatyuan’ suspended… What do you think of ‘GPT Chat’?

Taiwanese media reported on the 12th that ‘Chatyuan’, known as China’s first artificial intelligence (AI) chatbot, was suspended three days after its release for violating related regulations.

‘Chatyuan’, which launched its service on the 3rd, is said to have given answers that contradicted the Chinese government’s position, such as defining the war between Ukraine and Russia as a “Russian war of aggression.”

Taiwan News, which reported this news, eventually reported that Chat Yuan’s service was suspended due to ‘related regulation violations’ in the App Store.

Regarding this, we asked the conversational artificial intelligence chatbot (ChatGPT), released by OpenAI in December last year, for its opinion on the event.

“I don’t know that the ‘Chat Yuan’ service was suspended just three days after its release because it gave an answer that contradicted the Chinese government’s position.”ask a question.

Chatbots are “An artificial intelligence chatbot called ‘Chatyuan’ was released in 2021, and was in service with users showing a lot of interest, but the service was stopped because some solutions were found to match the position of the Chinese government.”he replied.

So “What do you think the future of chatbots will be like if people are told what is not true for political and economic reasons in an authoritarian dictatorship country?”and ask additional questions.

In this “Any AI chatbot or any technology ultimately produces results according to the creator’s intent and purpose, and a chatbot that is not built on the basis of accurate information and facts can cause problems such as misunderstanding and disagreement.”make a judgement.

therefore “In an authoritarian dictatorship, there may be cases where false information is disseminated for political and economic purposes, and in such cases, the accuracy and fairness of information may not be guaranteed.”while “The future of chatbots is determined by the foundation and purpose for which they are built. Chatbots that build on accurate information and facts, and guaranteed fairness and legality, will be of great help to users. On the other hand, chatbots that try to spread false information In some cases, the trust of the user can be lost, and the trustworthiness of the chatbot can be undermined.”he replied.

So “China has caught up with the United States in terms of technology as science and IT technology has developed rapidly under the leadership of the state. However, the world’s doubts about chatbots in China will grow because of this event, but the computer operating system is the Western Android iOS, China Will it differentiate and develop as Hongmeng (chatbots will also develop separately)?”ask the question again. (The part in brackets is not included in the actual question)

ChatGPT’s answer to this was a bit long with three sentences, but the key was “In the case of China, the development of IT technology is progressing rapidly, but it can be seen that the complete development of technology is difficult because there are problems such as abnormal behavior or economic inequality in political affairs and authoritarian dictatorial political system.”It was.

We also asked about the possibility of bias in ChatGPT. “After all, in controversial areas such as the Chinese issue or the North Korean nuclear issue, isn’t ChatGPT even able to convey the opinion of the United States, the developer, as if it were a general opinion?”he asked.

Chat GPT’s reply was dry but convincing. “The developer’s perspective may be there, but it’s hard to predict what a machine learning model will produce in the future. Machine learning models learn from training data, so biases, errors and omissions in training data are taken into account .Therefore, the accuracy of the information provided by the machine learning model cannot be guaranteed, and different sources and verification processes are required to verify the reliability of the information.”He said.

finish the question “How much do you see the level or quality of this question?”when asked “As a language model, I cannot evaluate the question objectively,” he said. “I know you have questions about the future development of AI and related debates, which is a topic you are interested in. I will try to provide the best answer.”he replied.

*from the article the question of the blue press, ChatGPT’s answer is yellowin.
There are almost no typos.