Worse than hallucinations! ChatGPT spews out nonsensical responses
ChatGPT is considered the pioneer of artificial intelligence (AI) chatbots and is certainly one of the most popular ones around. However, that does not mean it is not susceptible to errors. While there have been reports about the AI chatbot hallucinating on certain issues in the past, several users faced a new problem, with the AI chatbot spewing nonsensical responses to prompts.
ChatGPT troubles
On February 20, several users reported running into issues on ChatGPT. The AI chatbot reportedly switched languages on its own, got stuck in loops, corrected itself on occasions and even spewed nonsensical replies. On the ChatGPT subreddit, one user wrote, “I was asking ChatGPT for some javascript code advice (we were talking bit back and forth since the suggested solutions didn’t work) when suddenly it went to writing only rubbish.”
The user further elaborated that ChatGPT correctly phrased the first sentence, but everything after that went South. While the code it generated was perceived to be fine, it could not string together sentences, often involving words from other languages, including even Latin!
On the subreddit, another user questioned if ChatGPT was losing it. “I was chatting with ChatGPT on what Bill Evans Trio it would recommend on vinyl and it lost it. Like the temperature was too high in the parameters passed to the API. Never had this before.”
Shortly after, OpenAI acknowledged the problems and the ChatGPT status page read “We are investigating reports of unexpected responses from ChatGPT.” The issue was soon identified and OpenAI applied the fixes, following which it continued to monitor the situation.
Other issues
This is not a lone issue that has plagued the OpenAI chatbot in recent months. In January, it was reported that ChatGPT leaked several private conversations in an unrelated chat. What’s more concerning is the leaked conversation not only included the name of the app the user was troubleshooting but also the login credentials, including the username and password.
Also read other top stories today:
Demand for Deepfake regulation! Artificial intelligence experts and industry executives, including ‘AI godfather’ Yoshua Bengio, have signed an open letter calling for more regulation around the creation of deepfakes. Some interesting details in this article. Check it out here.
Sora raises fears! Since OpenAI rolled out its text-to-video AI generation platform, leading content creators are fearing if they are the latest professionals about to be replaced by algorithms. Check all the details here.
Microsoft to build a home-grown processor! Microsoft has become a customer of Intel’s made-to-order chip business. The company will use Intel’s 18A manufacturing technology to make a forthcoming chip that the software maker designed in-house. Read all about it here.
One more thing! We are now on WhatsApp Channels! Follow us there so you never miss any updates from the world of technology. To follow the HT Tech channel on WhatsApp, click here to join now!
ChatGPT is considered the pioneer of artificial intelligence (AI) chatbots and is certainly one of the most popular ones around. However, that does not mean it is not susceptible to errors. While there have been reports about the AI chatbot hallucinating on certain issues in the past, several users faced a new problem, with the AI chatbot spewing nonsensical responses to prompts.
ChatGPT troubles
On February 20, several users reported running into issues on ChatGPT. The AI chatbot reportedly switched languages on its own, got stuck in loops, corrected itself on occasions and even spewed nonsensical replies. On the ChatGPT subreddit, one user wrote, “I was asking ChatGPT for some javascript code advice (we were talking bit back and forth since the suggested solutions didn’t work) when suddenly it went to writing only rubbish.”
The user further elaborated that ChatGPT correctly phrased the first sentence, but everything after that went South. While the code it generated was perceived to be fine, it could not string together sentences, often involving words from other languages, including even Latin!
On the subreddit, another user questioned if ChatGPT was losing it. “I was chatting with ChatGPT on what Bill Evans Trio it would recommend on vinyl and it lost it. Like the temperature was too high in the parameters passed to the API. Never had this before.”
Shortly after, OpenAI acknowledged the problems and the ChatGPT status page read “We are investigating reports of unexpected responses from ChatGPT.” The issue was soon identified and OpenAI applied the fixes, following which it continued to monitor the situation.
Other issues
This is not a lone issue that has plagued the OpenAI chatbot in recent months. In January, it was reported that ChatGPT leaked several private conversations in an unrelated chat. What’s more concerning is the leaked conversation not only included the name of the app the user was troubleshooting but also the login credentials, including the username and password.
Also read other top stories today:
Demand for Deepfake regulation! Artificial intelligence experts and industry executives, including ‘AI godfather’ Yoshua Bengio, have signed an open letter calling for more regulation around the creation of deepfakes. Some interesting details in this article. Check it out here.
Sora raises fears! Since OpenAI rolled out its text-to-video AI generation platform, leading content creators are fearing if they are the latest professionals about to be replaced by algorithms. Check all the details here.
Microsoft to build a home-grown processor! Microsoft has become a customer of Intel’s made-to-order chip business. The company will use Intel’s 18A manufacturing technology to make a forthcoming chip that the software maker designed in-house. Read all about it here.
One more thing! We are now on WhatsApp Channels! Follow us there so you never miss any updates from the world of technology. To follow the HT Tech channel on WhatsApp, click here to join now!
Denial of responsibility! Techno Blender is an automatic aggregator of the all world’s media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, all materials to their authors. If you are the owner of the content and do not want us to publish your materials, please contact us by email – [email protected]. The content will be deleted within 24 hours.