Bangkok: Siriwong has responded to criticism regarding the TH-AI Passport and its handling of queries related to "blue shirt senators." He explained that when similar questions were posed to ChatGPT Pro, the AI provided a range of answers including "orange-shirt senator," "pink-Nanthana," and "blue-shirt Siriwong." Siriwong emphasized the variability of AI responses based on data and cautioned against blindly accepting AI outputs, labeling such actions as "absurdly foolish."
According to Thai News Agency, Mr. Siripong Angkasakulkiat, Deputy Minister of Transport and Deputy Leader of the Bhumjaithai Party, addressed the social media discussions that emerged after users engaged with the TH-AI Passport project through a word-color matching game. In the game, when users entered "banana," the system replied "yellow," and for "rambutan," it replied "red." However, when "senator" was entered, the system responded with "blue," leading to jokes and political implications.
Mr. Siripong shared his own experience with the ChatGPT system, where he tested a command prompting the AI to respond with a color for each word. The AI matched "banana" with yellow, "apple" with red, "senator" with orange, "Nanthana" with pink, and "Siripong" with blue. He commented on the perception that AI responses could influence intelligence but noted that he did not view it in such binary terms despite not being an expert.
He highlighted that many users have experimented with AI, citing an example of a navigation app that once led him to a dead end due to data processing errors. Mr. Siripong pointed out that these experiences can differ widely based on the data available and the user's location, and he observed that apps have improved over time with increased usage. He expressed that while TH-AI Passport and AI Pro might not provide definitive answers, they serve as tools for learning.
Mr. Siripong concluded by advocating for the correct use of AI, which involves asking questions, seeking answers, and making observations. He recalled the early days of AI when answers were often inaccurate, necessitating further inquiry. He believes that the TH-AI Passport can teach users to ask questions and verify answers scientifically. He warned against accepting AI responses without scrutiny, as it could lead to misinformation and ignorance.