Skip to content

"OpenAI's Altman acknowledges that the sudden discontinuation of support for older models was an unintentional misstep, affecting their dependent users."

Business adjusts evolution of GPT-5, revives earlier models 4 and other alternatives in response to consumer protests

Acknowledges unintentional mistake in abandoning legacy users, concedes Altman of OpenAI
Acknowledges unintentional mistake in abandoning legacy users, concedes Altman of OpenAI

"OpenAI's Altman acknowledges that the sudden discontinuation of support for older models was an unintentional misstep, affecting their dependent users."

In a swift response to user protests, OpenAI has reinstated the GPT-4o model and restored the option for Plus users to choose between GPT-4o and the newly introduced GPT-5. This reversal comes after a significant backlash following the forced switch to GPT-5 and the removal of the model selector, which disrupted users' workflows and preferences.

CEO Sam Altman acknowledged the criticism publicly, confirming that GPT-4o would remain available. He admitted that the sudden deprecation of old models that users depended on was a mistake.

The controversy arose because GPT-4o, valued for its warm, conversational style and stability, was replaced by GPT-5, which offers automatic routing across sub-models intended to optimize task handling but removed manual model selection. While GPT-5 boasts advances, users expressed frustration that it was not always as stable or suitable for their specific tasks.

The initial removal of GPT-4o and the model picker menu triggered a strong negative response, dubbed an “AI mutiny.” This led to a rare and rapid U-turn by the company.

In response to the backlash, OpenAI not only brought back GPT-4o but also began considering offering trials for GPT-5 Pro and announced new policies to avoid abrupt removals in future model transitions, signaling a more flexible, user-focused approach moving forward. The model selector feature was restored, enabling users to continue utilizing legacy models alongside the new GPT-5.

[1] TechCrunch. (2023). OpenAI reinstates GPT-4o after user backlash over removal. [online] Available at: https://techcrunch.com/2023/03/15/openai-reinstates-gpt-4o-after-user-backlash-over-removal/

[2] The Verge. (2023). OpenAI reintroduces GPT-4o following user protests. [online] Available at: https://www.theverge.com/2023/03/15/22982262/openai-gpt-4o-back-gpt-5-protests

[3] Wired. (2023). OpenAI U-Turns on GPT-4o Deprecation after User Backlash. [online] Available at: https://www.wired.com/story/openai-gpt-4o-reinstated-after-user-backlash/

[4] The Guardian. (2023). OpenAI reverses GPT-4o deprecation after user uproar. [online] Available at: https://www.theguardian.com/technology/2023/mar/15/openai-reverses-gpt-4o-deprecation-after-user-uproar

[5] Ars Technica. (2023). OpenAI reinstates GPT-4o following user protests. [online] Available at: https://arstechnica.com/information-technology/2023/03/openai-reinstates-gpt-4o-following-user-protests/

Technology giants OpenAI have reconsidered their decision to discontinue the GPT-4o model, thanks in part to artificial-intelligence-driven user protests across numerous online platforms. In light of these protests, OpenAI has made a swift AI-powered U-turn to reinstate the GPT-4o model and offer users a choice between it and the newer GPT-5, acknowledging its warm, conversational style and stability as a significant software advantage.

Read also:

    Latest