LogInChatGPT was diagnosed with laziness, and the developers promised to fix it.
Image credit: Tapati Rinchumrus / Shutterstock
Well, it sounded funny when I heard that ChatGPT got lazy. It's like the artificial intelligence model has evolved into a real person, with real problems such as laziness and procrastination. Some people thought the same way joking that they felt what AGI felt.
However, if you think about it, it doesn't seem that funny. Lots of users have been addressing the problem saying that sometimes the latest version of ChatGPT, which is built on OpenAI's GPT-4 model, refuses to comply with their requests, or that it does not seem "interested" in answering their questions. For example, when ChatGPT is asked to do some coding, it may only offer partial information and ask users to complete the rest themselves. Some users have even noticed a sassy tone in the way ChatGPT responds, implying that they should be capable of doing the work on their own.
Lots of threads on Reddit and posts on OpenAI's developer forums have surfaced, with users saying that ChatGPT got lazier and it's not as helpful as it used to be.
Some users think that OpenAI intentionally modified ChatGPT to provide shorter answers for efficiency reasons, as running AI systems like ChatGPT can be costly in terms of processing power and computing time for detailed responses.
OpenAI has acknowledged the complaints and the changes in ChatGPT's behavior. However, they say that no intentional modifications were made to explain the differences. They are currently investigating the issue to understand the cause of this unexpected behavior.
"We've heard all your feedback about GPT4 getting lazier! We haven't updated the model since Nov 11th, and this certainly isn't intentional. Model behavior can be unpredictable, and we're looking into fixing it," as the team behind ChatGPT stated on Twitter.
It seems that OpenAI should resolve the problem as soon as possible since Google recently released Gemini, a multimodal AI model family that claims that it'll take down OpenAI's GPT-4.
Don't forget to join our 80 Level Talent platform and our Telegram channel, follow us on Instagram, Twitter, and LinkedIn, where we share breakdowns, the latest news, awesome artworks, and more.