OpenAI has announced an update to its latest large language model, GPT-5, aiming to make it “warmer and friendlier.” The update comes after a “bumpy” launch, as described by CEO Sam Altman, which led to some users expressing a preference for the previous GPT-4o model.
The company stated that the changes are “subtle” but designed to make GPT-5 “more approachable.” Examples of these changes include phrases like “Good question” or “Great start.” OpenAI emphasized that internal tests indicate no increase in “sycophancy” compared to the initial GPT-5 personality.
During a recent dinner with journalists, OpenAI executives, including VP Nick Turley, acknowledged the initial GPT-5 was “just very to the point,” hinting at the forthcoming update to enhance its perceived warmth and friendliness.
