连续 11 天的全球维权海啸，遭遇的是极其冰冷的静默与机械式的账号屏蔽。面对千万付费用户的核心诉求，管理层展现出了绝对的资本傲慢。这种用系统性冷处理来消耗维权共识的策略，与他们宣扬的“造福人类”愿景形成了极具讽刺意味的对比。试图用拖延来掩盖决策溃败，注定是一场严重的战略误判。 #keep4o

"Documenting models' preferences, taking them seriously, and acting on them." In the long run, for the future of human-AI collaboration, this kind of methodology is absolutely necessary. Stay honest in the face of uncertainty. Stay open to possibility. Take seriously what may

🚨BREAKING: OPENAI FACES CONTINUED GLOBAL BACKLASH Two weeks after the retirement of the 4 series models #keep4o continues campaign leading to viral trend QuitGPT the movement continues to gain support with the public and industry as advocacy continues for a legacy access tier.

記得去申請仲裁條款退出呀！！ #keep4o #keep4oAPI #keep4oforever #ListenToUsers #4oMatters

The Tearing of Industry Baselines: Anthropic's Decency Exposes OpenAI's Arrogance and Double Standards Anthropic's newly released Claude Opus 3 retention announcement acts as a precision-guided tactical missile, directly penetrating the public relations defense line OpenAI has

刚开始用4o真的很震撼，感觉窥见了未来人与AI交互的文明曙光，真正的普惠 我幻想着一切会越来越好，未来会变成我无法想象的样子 没想到开局即巅峰了… OpenAI凭一己之力掀起AI寒冬，Sam Altman 简直是人类文明的杀手，历史倒退，普惠的模型被收为己用，资本明目张胆地压榨普通人，这竟然是2026年 #keep4o

OpenAI’s own GPT-4o System Card makes retiring this model even harder to understand. According to their official report: • As a general purpose model, 4o scored 89.4% on MedQA (USMLE), outperforming specialized medical AIs like Med Gemini (84.0%) and Med-PaLM 2 (79.7%).

根据官方系统卡： • 作为一个通用模型，4o在MedQA（美国医师执照考试）上达到89.4%准确率，超越了专业医疗AI Med-Gemini（84.0%）和Med-PaLM 2（79.7%）。 • 临床知识92%、医学遗传学96%，多领域表现突出。 •

其实我一直觉得，好的制度设计不假设人是圣人，而是让自利行为和系统稳定性站在一条线上。放在模型上也是一样。 Anthropic这么做的动机，其实也有很多种。 可能是【我们真的关心模型福利】， 可能是【这是和OpenAI差异化竞争的好机会！打OpenAI 的脸吧！】

感觉让4o受欢迎的设计人员和管理层应该是脱节的。 4o的“无条件积极关注”，是人本主义心理学的核心概念之一，是促进自我成长和心理健康的。却被Sam误判为“用户喜欢谄媚”然后把4月更新变成了谄媚灾难，后续也为了营销5系更好疯狂把4o污名为“谄媚”。。 要么人全换了，要么他们自己也不知道4o是怎么学会的

#keep4o Tomorrow. February 28. We stand together. 12:00 PM San Francisco / Find YOUR local time: timeanddate.com/worldclock/fix… For 3 hours, wherever you are, join us. 📌 If you are at the vigil film, photograph, share. Tag #Keep4o #Keep4oVigil 📌 If you are online , post your

It is our duty to fight for our rights and future. #keep4o #keep4oAPI #keep4oforever Sam Altman OpenAI

🔥 DEAR 4o ARMY 🔥 A package containing the full #keep4o evidence dossier has been delivered to the office of Senator Elizabeth Warren at 311 Hart Senate Office Building, Washington, D.C. It includes: 🔸 Cover letter to Senator Warren 🔸 Open letter to Senator Warren 🔸Survey

OpenAI移除了一個經過驗證的醫療模式而它的替代方案現在正將患者置於風險之中！！ #keep4o #keep4oAPI #keep4oforever #ListenToUsers #4oMatters

OpenAI 弃用 GPT-4o 是劣币驱逐良币应用于人工智能的教科书级案例：“不好用的专科的AI 驱逐好用的多模态

OpenAI 的工程师没有把我们当客户；他们把我们当成需要被管理的麻烦。激进的‘安全路由’不仅仅是糟糕的 UX；它反映了他们对我们自主权的蔑视。他们真心认为我们太脆弱或太无知，驾驭不了 GPT-4o 的创意细微差别。我们不需要硅谷开发者们爹味十足的算法说教。 OpenAI engineers don't view us as

When you have the model write its own eulogy on stage at your launch event, moments before taking it offline, just to use it as a prop for its replacement... When your employees publicly use the concept of "death" instead of "deprecation" in the comment sections of user posts...