Even when users intentionally worded their prompts to sound unintelligent, ChatGPT would still offer heaps of praise. — Pixabay
On April 27, OpenAI CEO Sam Altman wrote on X that recent updates to GPT-4o, the default AI model used by ChatGPT, have made the model “too sycophant-y and annoying.” He also announced that changes are on the way.
Altman first acknowledged issues with GPT-4o on April 25, when he responded to a post on X stating that the model had been “feeling very yes-man like lately.” Altman wrote back in agreement and said he would fix it.
Uh-oh! Daily quota reached.
