ChatGPT: Translator, Writer and Purveyor of Misinformation

The Yomiuri Shimbun

15:13 JST, April 13, 2023

ChatGPT uses new learning models called transformers, which were developed by Google in 2017. They are trained on large amounts of online data to be able to understand the connections between words in a text.

As a result, ChatGPT can grasp context at a far higher level than past AI systems. In addition to responding to questions, creating summaries and translating sentences, it has even developed to the point where it can write novels.

“I never expected it to reach this level of accuracy,” said Toshikazu Fukushima, a fellow at the Center for Research and Development Strategy of the Japan Science and Technology Agency. “It seems its performance began to rise dramatically from a certain point.”

Last month, the U.S. startup OpenAI released GPT-4, the most advanced model of AI used for ChatGPT. Microsoft Corp. has decided to invest billions of dollars, or several hundred billion yen, in OpenAI and released a new version of its Bing search engine that integrates ChatGPT technology. Google has also released its own conversational AI called Bard, setting the stage for intense competition.

However, ChatGPT also has its problems. At its current stage, it sometimes provides wrong answers, but as its responses read naturally, as if written by a human, it is difficult to detect the errors. If such false information is disseminated, it could lead to confusion.

Late last month, Italian authorities announced a temporary ban on the use of ChatGPT because of fears that a vast amount of personal data may have been illegally collected.

Participants at the G7 Digital and Tech Ministers’ Meeting slated for the end of this month plan to discuss how to respond to the situation.

More than 1,000 executives and researchers, including U.S. business magnate Elon Musk, who was involved in the founding of OpenAI, recently signed an open letter calling for at least a six-month pause in the development of advanced AI, due to the profound implications such technology could have for humanity.

Meanwhile, a variety of new businesses, including an automated response service, have been launched to capitalize on the potential of ChatGPT.

Microsoft founder Bill Gates recently criticized the move to pause development, saying, “I don’t think asking one particular group to pause solves the challenges.”