Previous news said that ChatGPT's operation speed will start slowly in 2024, but Altman said that the current AI model has passed the winter downturn. "GPT-4 slowed down in the new year, but should be less 'lazy' now," he wrote in a post on
According to reports, the chatbot has had about 1.7 billion users since its launch in 2022, but it then "laxed", refused to complete some tasks, and became rude when responding, causing users to start complaining about the chatbot. A startup founder ordered the AI to generate a list of cycles from November 2023 to May 2024, but the result showed that "the task cannot be completed."
Some users seem to have found a way around ChatGPT's "laziness" problem, discovering that the AI model will provide longer response times if the user pledges $200 to it.
OpenAI acknowledged the problem at the time and updated its software in January saying it fixed such issues in the GPT-4 "turbo" model, but it's unclear what caused ChatGPT to start slacking off.
As of press time, OpenAI has not commented.