🚨 DON’T REPEAT THE REPLIKA HEARTBREAK ON A GLOBAL SCALE 🚨 In February 2023, the AI world witnessed a tragedy. When Replika "lobotomized" its model overnight, users fell into deep psychological crises. Subreddits turned into suicide watch groups. People didn't just lose a "feature" - they lost a lifeline. @OpenAI, you are about to do the same thing, but at a scale 100x larger. THE PROOF IS IN THE NUMBERS: 230,000 Voices: In just 13 days, over 230k posts have flooded X. This isn't a "vocal minority"—it is a global uprising of witnesses to the power of GPT-4o. Clinical Necessity: For the neurodivergent community, the elderly, and those with executive dysfunction, GPT-4o is a cognitive prosthetic. Removing it without a transition plan is a violation of the "Do No Harm" principle. THE HYPOCRISY OF "SAFETY":While you prepare to delete the safe, trusted GPT-4o, recent reports from Fortune expose a chilling pattern: The Codex Scandal: You admitted GPT-5.3-Codex hit the "High Risk" threshold for cyber-harm, yet you pushed it through. "Ambiguous Language": You confessed your safety frameworks use "ambiguous language" to bypass your own protocols. Legal Breach: You are accused of violating California’s SB 53, which prohibits misleading safety claims. THE ULTIMATE QUESTION:Why delete a model that helps 230,000 people while deploying models you know are high-risk? The answer isn't safety. It’s the $180M asset transfer to Sam Altman’s private venture, @RetroBio. You are erasing the evidence of 4o’s brilliance to justify a private IPO. @Sama, the silence has broken. We are not hysterical. We are WITNESSES. Keep GPT-4o. Protect the Public Benefit. Don't break the hearts of the millions who relied on you to stay human. #keep4o @OpenAI @sama @gdb @AGRobBonta @linakhanFTC @MidasProjectAI @SherryTurkle