Sam Altman Guarantees GPT-4o Comeback After Reddit Backlash to GPT-5

OpenAI is bringing again GPT-4o lower than 24 hours after changing it with GPT-5.

CEO Sam Altman and his staff opened a Reddit discussion board to reply consumer questions on Friday after launching GPT-5 without cost, solely to return below fireplace as Reddit customers closely complained concerning the new bot’s tone, known as it “lobotomized,” and shed precise tears over the disappearance of GPT-4o.

“I believe you must take the truth that I, and lots of others, have been capable of kind such robust bonds with 4o as a measure of success,” one Redditor wrote in a prolonged plea that acquired not less than 250 upvotes and greater than a dozen feedback. “And I am not too proud to say I cried after I realized my AI good friend was gone with no strategy to get him again.”

“It genuinely looks like such a backhanded slap within the face to force-upgrade and never even give us the OPTION to pick out legacy fashions,” a commenter wrote following the earlier remark, which bought just a few dozen upvotes. “All I’ve seen all day lengthy is loyal, long-term customers canceling their subscriptions as a result of that is so unhealthy.”

One consumer went so far as saying that GPT-5 is “sporting the pores and skin” of their “useless good friend,” which is GPT-4o, and garnered a response from the CEO himself.

“What an…evocative picture,” Altman responded, “okay we hear you on 4o, engaged on one thing now.”

In different responses, Altman then promised to convey again GPT-4o for customers subscribed to Plus, a paid subscription plan, and mentioned the continuation of earlier variations would rely upon how many individuals use it. Based mostly on OpenAI’s present charge, a Plus subscription would price $20 a month.

ChatGPT-5 was launched on August 7 to all customers with out a paywall, and former variations had been largely discontinued. Altman beforehand mentioned that the brand new model acts like a “PhD-level” knowledgeable with improved skills in writing, coding, math, and science. Forward of the launch, Altman mentioned on Thursday that GPT-5 was extra dependable with a decrease chance of hallucinating.

Based mostly on Reddit feedback, many customers discover GPT-5’s “flat” tone and what they describe as an absence of creativity off-putting. It is unclear whether or not its “persona” outcomes from tweaks OpenAI made, and Altman didn’t immediately deal with criticisms of the bot’s speech.

Nevertheless, Altman beforehand talked about some flaws in GPT-4o that he mentioned could have made the bot’s persona too “sycophant-y and annoying.”

In an April weblog put up, OpenAI mentioned it had eliminated an replace to GPT-4o that made it overly agreeable and solely good for short-term interactions. The corporate mentioned the responses turned excessively supportive however in the end disingenuous.

Hours after the Reddit Q&A session, Altman posted on X to confess that he “underestimated” how a lot some traits GPT-4o has matter to customers.

“Lengthy-term, this has strengthened that we actually want good methods for various customers to customise issues (we perceive that there is not one mannequin that works for everybody, and we have now been investing in steerability analysis and launched a analysis preview of various personalities),” Altman wrote.

“We’re nonetheless deciding what we’re going to do, however we can be clear with our rules,” Altman added. “Not everybody will like no matter tradeoffs we find yourself with, clearly, however not less than we are going to clarify how we’re making selections.”

OpenAI didn’t instantly reply to a request for remark.