You are not alone if you've been finding ChatGPT and GPT-4 to be noticeably lazier than before. Its maker, OpenAI, has acknowledged this problem and said it is working on a fix.
What Happened: The Microsoft Corp.-backed (NASDAQ: MSFT) OpenAI has issued a statement acknowledging a problem that many ChatGPT users have been complaining about for several days now.
"We've heard all your feedback about GPT4 getting lazier! We haven't updated the model since Nov 11, and this certainly isn't intentional," OpenAI said in a post.
However, the company said AI models could be "unpredictable." While this does not suggest that GPT-4 is out of OpenAI's control, it means that it's not possible to predict 100% of the outcomes or paths that GPT-4 can take to answer a given question.
Essentially, some blame for ChatGPT and GPT-4 getting lazier could be due to a scenario that OpenAI has not addressed or has missed while creating the model.
Several users on social media have been complaining about GPT-4 getting worse. Instead of doing the task that users have asked it to, the AI chatbot returns with a sample solution and asks the users to finish the rest of the job themselves.
It's not very polite of an AI chatbot to disregard instructions, is it?
ChatGPT Has Not Become Sentient Yet: If you're worried that ChatGPT has become sentient and is now beginning to behave like its creators - humans - then you can shrug them off.
"To be clear, the idea is not that the model has somehow changed itself since Nov. 11. It's just that differences in model behavior can be subtle," OpenAI said.
It clarified that not every subset of issues may be immediately predictable or noticed by users or employees.
For now, though, we'll have to deal with a lazier ChatGPT.