Artificial Intelligence

ChatGPT gave odd responses to many users; OpenAI acknowledges the issue

ChatGPT users started noticing strange responses from the chatbot recently. These included switching between languages, getting stuck in loops, or even repeatedly correcting itself. Some of the responses were just plain gibberish.

When talking about the Jackson family of musicians, the chatbot told a Reddit user, “Schwittendly, the sparkle of tourmar on the crest has as much to do with the golver of the ‘moon paths’ as it shifts from follow.”

At 10:40 PM ET, OpenAI admitted on its status page that it was “investigating reports of unexpected responses from ChatGPT,” and a few minutes later, they said they had identified the problem and were working on a solution.

As of now, the status page still has a yellow-bordered box at the top saying the company is “still keeping an eye on the situation.” Many of the errors seem to resemble a popular social media meme where people are asked to type a word on their phone and then repeatedly tap the same next word suggestion to see what happens. This makes sense because large language models are essentially just very advanced predictive text.

Take these examples with a grain of salt — it’s quite simple for trolls to create their own images like this, after all. However, there are some recurring themes in social media posts discussing the problem, such as ChatGPT getting stuck in a loop. In the images above, for instance, ChatGPT seemed to repeat “.Happy Listening!

Reddit users have experienced similar problems before. We’ve contacted OpenAI to get more details about the issue. It’s a good reminder that generative AI chatbots are a rapidly evolving technology. Moreover, the large language models that drive them are essentially black boxes, even to their creators, and are not fully understood in a highly detailed manner.

They work by identifying patterns in data and then making their best guesses about how to respond to input, which often results in impressive outcomes. However, it can also lead to hallucinations, emotional manipulation, and even laziness.

Rohan Sharma

Recent Posts

Best Video Editing Software For PC

Video editing is one of the most in-demand skills in today’s content creation era. If…

7 months ago

Samsung planning to introduce blood glucose monitoring with Galaxy Watch 7

There have been whispers about Samsung's ambition to equip their wearable gadgets with a neat trick:…

7 months ago

TSMC to lock horns with Intel with its A16 chip manufacturing tech

Taiwan Semiconductor Manufacturing Co (TSMC) recently dropped the news that they're gearing up to kick off production…

7 months ago

Is ChatGPT accurate and should we believe what it says?

Modern chatbots like ChatGPT can churn out dozens of words per second, making them incredibly…

7 months ago

Mark Zuckerberg claims Meta is years away from making money through gen AI

The race for generative AI is in full swing, but don't count on it raking…

7 months ago

How JioCinema’s dirt cheap plans can mean trouble for Netflix, Amazon Prime

JioCinema, the famous Indian on-demand video-streaming service, unveiled a new monthly subscription plan, starting at…

7 months ago