A growing coalition of people is expressing frustration over GPT-5 Proβs failure to follow instructions, marking a significant shift for individuals relying on AI for coding and planning. Complaints have surged in recent days, with many questioning the system's dependability.
Concerns surrounding GPT-5's reliability are mounting. Users have reported repeated failures to follow basic directives. One user stated, "I tell it 'donβt give me code,' yet it responds with pages of code anyway." Others share similar experiences, highlighting that the AI often produces irrelevant responses, losing track of context.
Interestingly, a new comment indicated, "They started to route your requests to different models to test them." This could imply that current issues may stem from the AI's ongoing adjustments. Another user suggested canceling subscriptions to express discontent over being viewed as beta testers.
Reports now suggest that many users notice:
Memory Issues: Users observe the AI quickly forgetting prior inputs, hindering continuity.
Response Inconsistency: One user detailed their struggle, stating they repeatedly asked the AI to change image dimensions without success.
Personalization Decline: Feedback indicates a reduction in the AI's ability to cater to individual needs since previous versions like GPT-4.
As frustration peaks, thereβs a growing desire among users to revert to older AI models, with many stating that GPT-4 provided a more robust experience.
Overall feedback leans heavily negative. Key sentiments include:
Ignoring Previous Inputs: An increasing number of users report that the AI does not recognize earlier inputs, leading to redundant prompts. "It needs two prompts just to reflect on a previous point,β expressed one individual.
Frustrating Dynamics: Disillusionment seems prevalent, with complaints like, "I feel like this is a joke; thereβs no option to use previous versions."
Exploring Alternatives: As dissatisfaction grows, more individuals are investigating other AI options. "Itβs time to look at alternatives," echoed one voice from the forums.
With persistent criticism, developers may need to prioritize enhancements for GPT-5βs functionality. Industry experts suggest a 60% chance of significant improvements to memory and context retention in upcoming months. Will this reshape user interactions with AI?
β‘οΈ 75% of comments highlighted forgotten context and irrelevant responses.
π Many seek a return to older models, emphasizing improved reliability.
π‘ Users report a substantial decline in personal interaction quality with the AI.
As community discussions grow stronger, it will be interesting to see how these sentiments might push for changes in AI developments, or encourage users to pursue alternative technologies.