OpenAI responded to the complaints that ChatGPT's GPT-4 is getting "lazier." The AI company responded on X saying, "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."
In short, OpenAI is saying they have not made any recent changes to ChatGPT or GPT-4 that would make it lazier. In fact, no changes have been made since November 11th. But they explained that thee model itself "can be unpredictable" and that is what they are looking into fix.
OpenAI added, "to be clear, the idea is not that the model has somehow changed itself since Nov 11th. it's just that differences in model behavior can be subtle -- only a subset of prompts may be degraded, and it may take a long time for customers and employees to notice and fix these patterns."
Here are those posts:
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 🫡
— ChatGPT (@ChatGPTapp) December 8, 2023
to be clear, the idea is not that the model has somehow changed itself since Nov 11th. it's just that differences in model behavior can be subtle -- only a subset of prompts may be degraded, and it may take a long time for customers and employees to notice and fix these patterns
— ChatGPT (@ChatGPTapp) December 8, 2023
Meanwhile, they also posted this over the weekend:
when releasing a new model we do thorough testing both on offline evaluation metrics and online A/B tests. after receiving all these results, we try to make a data driven decision on whether the new model is an improvement over the previous one for real users
— ChatGPT (@ChatGPTapp) December 9, 2023
we’re always striving to make our models more capable and useful for everybody across millions of use cases. so please keep the feedback coming! it helps us stay on top of this dynamic evaluation problem 🙏
— ChatGPT (@ChatGPTapp) December 9, 2023
Forum discussion at X.
