Next Article
OpenAI is updating GPT-5 to make it friendlier
OpenAI is updating GPT-5 just weeks after launch, following feedback that it felt less friendly and more robotic than GPT-4o.
To make conversations feel warmer, the team is adding small touches like supportive phrases ("Good question") and new control modes—Auto, Fast, and Thinking—giving users more control over how GPT-5 operates.
The goal is to strike a balance between human touch
GPT-5 will keep its strengths in coding, math, writing, and science.
OpenAI is also tweaking rate and context limits to improve the overall experience based on what users want.
The goal: smarter chats that still feel human—without going back to GPT-4o's overly agreeable style.