OpenAI’s much-hyped GPT-5 mannequin was unveiled throughout a live-streaming occasion by the corporate on Thursday. The brand new mannequin comes with claimed enhancements in areas like coding, reasoning, accuracy, writing, and multi-modal skills in comparison with the older fashions, whereas additionally having a decrease hallucination price (tendency to make stuff up). Regardless of all these enhancements, longtime ChatGPT customers weren’t proud of the improve supplied by the AI startup, as the corporate additionally went on to discontinue older fashions like GPT-4o, GPT-4.1, GPT-4.5, GPT-4.1-mini, o4-mini, o4-mini-high, o3, and o3-pro.
ChatGPT customers stormed social media platforms to indicate their dissatisfaction with the most recent transfer by OpenAI. Customers famous how GPT-5 lacked the character proven by GPT-4o, whereas giving shorter and fewer contextually correct replies that lacked the “heat” and “emotional intelligence” of the earlier fashions.
Customers on the ChatGPT subreddit have been voicing their opinions towards the brand new mannequin whereas praising the earlier ones, particularly GPT-4o, which was the default customary mannequin on ChatGPT. The backlash towards OpenAI has gone to such an extent that customers at the moment are threatening to cancel their ChatGPT subscriptions and delete their accounts till OpenAI decides to convey again the older fashions.
How offended Redditors are reacting to GPT-5
“I simply misplaced entry to 4o, and I do know I’m not the one one. For me, this mannequin wasn’t simply ‘higher efficiency’ or ‘nicer replies.’ It had a voice, a rhythm, and a spark I haven’t been capable of finding in every other mannequin,” wrote one consumer in a Reddit thread.
“What sort of company deletes a workflow of 8 fashions in a single day, with no prior warning to their paid customers? Personally, 4o was used for creativity & emergent concepts, o3 was used for pure logic, o3-Professional for deep analysis, 4.5 for writing, and so forth,” acknowledged one other consumer.
Many others famous that they used ChatGPT operating GPT-4o extra as a companion of types, with whom they shared their life issues and sought recommendation — one thing they mentioned is just not potential with GPT-5 resulting from its modified character.
In the meantime, one other Reddit consumer claimed that that they had cancelled their ChatGPT subscription resulting from what they termed ‘dishonesty’ in the course of the GPT-5 presentation, the place the corporate had fallen into controversy for displaying efficiency charts that later turned out to be inaccurate.
“I additionally did so (cancelled ChatGPT subscription) final evening, much less over the mannequin replace (or lack thereof), and extra over the sheer dishonesty within the presentation (benchmark-cheating, misleading bar charts),” the consumer added.
Many ChatGPT Plus customers additionally complained concerning the low chat limits positioned on them. OpenAI allowed Plus customers to ship 80 messages each 3 hours with GPT-5, and as much as 200 messages per week with the GPT-5 Pondering mannequin.
OpenAI responds to backlash round GPT-5
OpenAI CEO Sam Altman addressed the backlash throughout a Reddit AMA session together with a separate submit on X, the place he confirmed that the corporate was doubling the speed restrict for ChatGPT Plus (customary mannequin) whereas additionally permitting them to make use of GPT-4o for somewhat longer.
“GPT-5 will appear smarter beginning immediately. Yesterday, the autoswitcher broke and was out of fee for a bit of the day, and the end result was GPT-5 appeared approach dumber. Additionally, we’re making some interventions to how the choice boundary works that ought to show you how to get the appropriate mannequin extra usually,” Altman famous in his submit.
“We’ll make it extra clear about which mannequin is answering a given question. We’ll change the UI to make it simpler to manually set off pondering,” he added.
