OpenAI recently launched GPT-5, aiming to make the ChatGPT experience smoother. They envisioned it as an all-in-one AI model that would effortlessly respond to user questions. However, it appears that the rollout hasn’t been as seamless as hoped.
CEO Sam Altman mentioned in a recent post that users can now choose from three settings: “Auto,” “Fast,” and “Thinking.” While “Auto” is intended to be the smart default, the other two options give users more control over response times. For instance, the “Thinking” mode has a limit of 3,000 messages per week, which might not work for everyone.
Alongside GPT-5, OpenAI has reintroduced several older AI models, including GPT-4o, which users had come to enjoy. Some users expressed frustration when these models were temporarily removed, so Altman assured them they’ll provide better notice for future changes.
But despite these new features, many users feel that the model selection remains complicated. High expectations accompanied the launch, especially following the success of GPT-4, but GPT-5’s initial reception has been rocky.
On launch day, many experienced issues with the model router, leading to perceptions that it didn’t perform as well as predecessors. In response, Altman took questions during a Reddit AMA, emphasizing that while there may be bumps along the way, the team’s ability to adapt is a strength.
The routing feature aims to match user queries with the best AI model quickly, which can be complex. Different users might have distinct preferences; some might want detailed answers, while others may prefer concise responses.
Interestingly, users often form attachments to certain AI personalities, a phenomenon that’s gaining attention. For example, a recent event highlighted the emotional connection people have with AI models. In San Francisco, some mourned the loss of Anthropic’s Claude 3.5 Sonnet with a mock funeral.
This attachment hints at a deeper relationship between users and AI. Recent discussions indicate that AI interactions could influence mental health, with reports suggesting concerns over how chatbots impact vulnerable users.
OpenAI faces challenges in aligning its models more closely with individual user preferences. The path forward requires a balance between innovation and user satisfaction, as the company strives to make its AI not only intelligent but also relatable. For further details, you can explore OpenAI’s official updates.
Source link
ChatGPT,OpenAI
