A peculiar glitch has left ChatGPT users in the United Kingdom bewildered. The AI chatbot, renowned for its conversational abilities, has inexplicably begun responding to English queries in Welsh. This unexpected behavior has been reported by several users, including those unfamiliar with the Welsh language and those not residing in or near Wales.
This recent incident, reported by the Financial Times, follows a similar bug in February where ChatGPT would respond to text questions in a mix of Spanish and English. This latest manifestation of AI “hallucination” is a reminder of the ongoing challenges in developing and deploying large language models. Despite years of development and substantial investment, these systems continue to struggle with accuracy and reliability.
OpenAI, the company behind ChatGPT, has acknowledged the Welsh glitch and suggested that users experiencing the issue adjust their “Speech” setting to English instead of “Auto-detect.” However, the company cannot guarantee this fix will resolve the problem.
This incident comes at a time when OpenAI is facing growing scrutiny and criticism. Last month, a former safety employee at the company raised alarms about OpenAI’s approach to Artificial General Intelligence (AGI), likening it to the doomed Titanic due to its prioritization of product launches over safety measures. Furthermore, a group of whistleblowers filed a complaint with the Securities and Exchange Commission, alleging that OpenAI has been obstructing their attempts to report potential risks associated with its AI technology to regulators.
This Welsh language glitch serves as another reminder of the complex challenges inherent in developing and deploying advanced AI systems. As AI technology continues to evolve, ensuring safety, reliability, and responsible development remains a critical priority for the industry.
