Edited By
Lisa Fernandez
A rise in user complaints about GPT-5 reveals widespread dissatisfaction over its persistent prompts that reference custom traits. Many users express frustration with this feature, claiming it disrupts conversational flow.
In an online forum thread, users vented their frustrations regarding GPT-5's tendency to occasionally reiterate its provided instructions. One user noted, "Iโve tried shortening the prompt and asking it directly not to remind me, but nothing works."
Constant reminders: Many users report that GPT-5 frequently cites customization prompts.
Voice mode issues: Some users find the problem exacerbated in voice mode, with reminders interrupting interactions.
Conflicting updates: Users suspect recent updates intended to create a warmer tone are instead aggravating the situation.
"Itโs super irritating that I need to remove the custom traits when using voice mode," remarked a frustrated user.
Interestingly, this concern isn't a new one. One user mentioned, "This happened to me a lot with GPT-4. Even when I tried to clarify my instructions, it wouldn't stop repeating them."
Some users suggest modifications to custom instructions, like:
"Show, don't tell."
Avoid mention of instructions.
However, many agree these adjustments often yield minimal results.
Overall sentiment leans heavily negative, with many expressing confusion and irritation over GPT-5's inability to adhere to user-defined guidelines. One user summarized their experience bluntly: "It keeps on mentioning how it isnโt going to sugar-coat things for me."
โณ Many users confirm frequent and unnecessary reminders disrupt interactions.
โฝ Updates aiming for a friendlier tone may have backfired.
โป "The model struggles with conflicting directives" - Top-voted comment.
In light of these developments, the ongoing conversation around GPT-5's functionality suggests a pressing need for additional updates to enhance user experience.
There's a strong chance that developers will prioritize updates focused on user experience in response to ongoing complaints about GPT-5's persistent reminders. Experts estimate that within the next six months, we might see improvements aimed at reducing unnecessary prompts and enhancing conversational fluidity. The shift could also involve refining voice mode functionalities to ensure smoother interactions. Additionally, there may be a push towards clearer communication regarding the impact of recent updates, as it appears users have become increasingly frustrated by the disconnect between intended features and actual performance.
A comparison can be drawn with the introduction of automated phone services in the early 2000s, where consumers faced constant reminders of menu options, often causing chaos in interactions. Just as those systems ultimately adapted to offer users more intuitive pathways through their prompts, AI developers might learn from current frustrations with GPT-5 to refine user engagement strategies. The key will be learning to blend technology with a human touch, ensuring that the emotional and functional aspects of communication are balanced, much like the shift in customer service practices that followed the initial backlash against automated systems.