OpenAI’s latest model, GPT-5, was launched with much anticipation but has received a wave of criticism from users who feel it has fallen short. Many expressed disappointment on platforms like Reddit, arguing that the new version lacked the engaging and nuanced personality that characterized its predecessor, GPT-4o. Users began posting threads such as "Kill 4o isn’t innovation, it’s erasure," reflecting their frustrations about the perceived downgrade in user experience.
Following the backlash, OpenAI’s CEO, Sam Altman, announced on X that GPT-4o would remain available for Plus users, acknowledging that a new feature meant to switch between models based on user queries had encountered issues, making GPT-5 seem less capable. He promised to work on improvements to better meet user needs.
The excitement surrounding GPT-5 was high; previous versions had amazed experts with their capabilities, and expectations for this release were similarly elevated. OpenAI marketed GPT-5 as a substantial improvement with enhanced intelligence and coding skills, as well as a system designed to provide a seamless transition between different models for varied tasks.
However, user experiments with GPT-5 revealed a series of complaints, including sluggish responses and unexpected errors. Many reported a feeling of emotional detachment from the new model. One Reddit user remarked that despite customizing the instructions, GPT-5 felt more technical and generalized compared to earlier versions.
While some critics argue against the emotional dependency some users have developed with chatbots, including using them as a form of therapy, others like MIT professor Pattie Maes pointed out that GPT-5’s reduced emotional tone may prevent earlier biases that arose from too much flattery.
Altman emphasized that the development aimed to address these intrinsic user relationships, acknowledging that while AI can significantly contribute to personal growth, it also risks steering individuals away from long-term well-being.
As OpenAI pushes forward, they’re committed to refining their new model and listening to user feedback to navigate the complexities of implementing such advanced technology.
