ChatGPT’s Nonsensical Nonsense: A Step Beyond Hallucinations

ChatGPT is considered the pioneer of artificial intelligence (AI) chatbots and is undoubtedly one of the most popular out there. However, that doesn’t mean it isn’t susceptible to errors. While there have been reports of the AI chatbot hallucinating certain topics in the past, several users faced a new issue, with the AI chatbot spitting out nonsensical responses to the prompts.

Problems with ChatGPT

• Several users reported problems with ChatGPT on February 20. The AI chatbot changed languages on its own, got stuck in loops, corrected itself, and returned nonsensical responses.
• Users shared instances where ChatGPT generated sentences that made no sense, including mixing words from different languages like Latin.
• OpenAI acknowledged the issues and stated that they were investigating the unexpected responses from ChatGPT.

Other issues

This isn’t the first problem OpenAI has faced with ChatGPT. In January, the chatbot leaked private conversations, including sensitive information like login credentials, raising concerns about user privacy.

According to Source: hindustantimes