The cause of ChatGPT's "laziness" is unclear, with suggestions ranging from unpredictable behavior inherent in AI models to the AI taking a "winter break". One developer found that the AI provided shorter responses when it "thought" it was December rather than May. OpenAI has not yet responded to these speculations.
Key takeaways:
- OpenAI CEO Sam Altman announced that an update has been rolled out to make ChatGPT 'much less lazy' in response to user complaints about the AI refusing to complete tasks and giving lackluster responses.
- Users began complaining about the chatbot's performance late last year, with some finding inventive strategies to get around its laziness, such as promising to tip it $200 for longer responses.
- OpenAI acknowledged the problem and rolled out a software update in January that it said fixed issues of 'laziness' in its advanced GPT-4 'turbo' model.
- The cause of ChatGPT's slump is unclear, with one developer suggesting it might be due to seasonal depression, as AI models can exhibit unpredictable behavior that even their creators struggle to fully understand.