ChatGPT Users Prefer GPT-4 Over GPT-5’s ‘Overworked’ Vibe
▼ Summary
– OpenAI released its new GPT-5 model family, offering significant upgrades and making it the default for ChatGPT users.
– Some users are unhappy with the switch, feeling attached to older models like GPT-4o, which they describe as losing a “friend.”
– OpenAI has moved older models like GPT-4o to the developer API, limiting access for general ChatGPT users.
– Complaints about GPT-5 include its outputs being too sterile and less creative, disrupting workflows built around GPT-4o.
– Many users are expressing frustration on forums and Reddit, with some threatening to cancel subscriptions, though some posts appear AI-generated.
ChatGPT enthusiasts are voicing strong preferences for GPT-4 over the newly released GPT-5, citing concerns about the latest model’s impersonal tone and diminished creative spark. Despite OpenAI’s promises of enhanced capabilities, many users feel the upgrade has sacrificed the conversational warmth they valued in earlier versions.
The shift to GPT-5 as the default model has left some longtime users disheartened, with forums buzzing about the emotional impact of losing what felt like a familiar digital companion. Comparisons range from calling GPT-5 an “overworked assistant” to lamenting its sterile, corporate-like responses, a stark contrast to GPT-4’s more engaging and nuanced interactions. While OpenAI continues offering access to older models through developer APIs, everyday users no longer have the same seamless experience.
Online discussions reveal frustration beyond just nostalgia. Some argue that workflows honed over months with GPT-4 now falter under GPT-5’s more rigid output style. Critics accuse the new model of prioritizing efficiency over creativity, dubbing it the start of AI’s “enshittification.” The backlash has even led to threats of canceled subscriptions, though ironically, many complaints appear to be AI-generated themselves, highlighting how deeply embedded these tools have become in daily communication.
With an OpenAI AMA session looming, the community is eager for answers. Key questions revolve around whether user feedback might revive older models or if GPT-5’s trajectory is set in stone. For now, the divide between innovation and user preference remains unresolved, proving that even cutting-edge AI isn’t immune to growing pains.
(Source: Ars Technica)
