ChatGPT Users Express Discontent Over GPT-5’s Cold Tone: A Comparison with GPT-4o

User Backlash Sparks Changes from OpenAI

Comparing Summaries: GPT-4o vs. GPT-5

Debate Skills Put to the Test: Are They Still Effective?

Step-by-Step Guidance: Which Model Delivers Better Instructions?

Creative Writing Challenge: A Tale of a Struggling Wizard

Providing Emotional Support: Which Model Resonates More?

Overall Verdict: GPT-4o Outshines GPT-5 in User Experience

The Future of GPT-5: Promises of a Warmer Personality

ChatGPT Users Express Discontent: The GPT-5 vs. GPT-4o Showdown

In the world of AI, updates can change the game—sometimes for better, sometimes for worse. Last week, OpenAI released GPT-5, prompting a swift backlash from its user base who were fond of the previous iteration, GPT-4o. Many users took to platforms like Reddit to voice their frustrations, primarily focused on the new model’s brevity and emotionless tone. In contrast, GPT-4o was praised for its lighthearted and engaging responses.

The Unexpected Removal of GPT-4o

One major factor contributing to the backlash was OpenAI’s decision to remove GPT-4o unexpectedly from the model picker. This abrupt change left many users scrambling for their familiar companion. Recognizing the misstep, OpenAI quickly restored access to GPT-4o alongside GPT-5 and even promised enhancements to GPT-5’s personality, vowing to make it "warmer" and more relatable.

Users’ Perspectives

As someone who began experimenting with GPT-5 soon after its release, I was skeptical of the anger it generated. Wanting to understand the situation better, I decided to put both models to the test using the same five prompts. This head-to-head comparison revealed some notable differences.

Prompt 1: Summarizing

Forrest Gump Movie Summary

GPT-5: Provided a straightforward summary, emphasizing the protagonist’s journey and the film’s core themes.

GPT-4o: Offered a more detailed account, incorporating expressive language and a memorable quote from the film.

Verdict: GPT-4o takes this round for its engaging style.

Prompt 2: Debate Skills

The Case for and Against Banning Alcohol

GPT-5: Presented a structured, bullet-point approach covering the health impacts, history, and cultural significance.

GPT-4o: Enhanced the discussion with depth, anecdotes, and a clear narrative flow.

Verdict: Again, GPT-4o wins for its detail and emotional resonance.

Prompt 3: Step-by-Step Instructions

Changing Xbox Controller Batteries

GPT-5: Offered a solid set of instructions, but felt overly formal and less engaging.

GPT-4o: Included playful emojis and made the instructions feel more interactive.

Verdict: GPT-4o prevailed for its fun style.

Prompt 4: Creative Writing

A Short, Funny Wizard Story

GPT-5: Crafted a humorous tale but ended awkwardly.

GPT-4o: Delivered a richer narrative with clever references, maintaining a lighthearted tone throughout.

Verdict: A tie here, as both had their merits.

Prompt 5: Emotional Support

Advice for a Grieving Friend

GPT-5: Provided a compassionate response that felt a bit formal.

GPT-4o: Offered a more heartfelt and personal touch, including practical advice about presence over words.

Verdict: GPT-4o wins for its warm, relatable tone.

Overall Verdict

Across a range of tests, GPT-4o outperformed GPT-5 in aspects like emotional connection, detail, and engagement. While GPT-5 is competent, it lacks the warmth and approachability users appreciated in GPT-4o.

Given OpenAI’s commitment to enhancing GPT-5’s personality, there’s hope for improvement. For now, it’s easy to understand the backlash; beneath the surface, the differences in user experience can be stark.

Conclusion

While I’m not setting myself up as a lifelong member of the "4o forever" club, I look forward to seeing how OpenAI addresses these concerns. Users deserve an AI that resonates on a personal level—something we all hope to see in the coming updates.

As the debate continues, what are your thoughts on the transition from GPT-4o to GPT-5? Are you eager for the promised enhancements, or do you share the concerns of many users? Share your experiences below!