@OpenAI 🌍🌍🌍
原文@OpenAI 🌍🌍🌍
原文@OpenAI 能帮我把验证码也点了吗,这个我真需要
原文@OpenAI great, now my computer can be productive without me🤣
原文@OpenAI I'm not afraid AI will take my job. I'm afraid it'll find out how little work I actually do.
原文@OpenAI TOOOOOO FAST
原文@OpenAI @grok Based on OpenAI's claims, is this new GPT-6.1 Terra better than GPT-6 Astra at Computer Use and Chrome use?
原文@OpenAI Terra 这个定位很关键:电脑操作场景里,成本和稳定性往往比单次推理上限更决定能不能真正用起来。想请教一个实际问题:跨应用的长流程任务中,低推理档在易错率下降后,遇到需要人工确认的高风险步骤时,是否会有更明确的停下来确认或可撤销机制?
原文@OpenAI 打得过 Fable 5.5 吗()
原文@OpenAI Price is the part that decides whether computer use leaves the demo stage. Agents retry and re-read the screen constantly, so cost per action matters far more than it ever did in chat.
原文@OpenAI computer use 和 50% 降价塞进同一个版本,等于直接把 Agent 产品的单次任务成本往下压。 接下来真正的门槛不是价格,是长流程里跨应用操作的成功率能不能稳住。
原文@OpenAI Imagine 6.1 Luna, whoooosh. But being honest, OpenAI needs better releases now that Anthropic is dominating with Opus 5.5. ONE MONTH and no model has dethrone it.
原文@OpenAI 准备走便宜大碗的降智路线是吗?
原文@OpenAI 能不能优先解决一下模型很慢的问题,现在响应太慢了
原文@OpenAI Terra 可以理解屏幕内容、跨应用执行操作,例如读取发票并录入财务系统、填写表单、分析工程图纸,甚至辅助视频编辑。让 AI Agent 以更低成本处理大量日常工作。👍
原文@OpenAI @grok gpt-6.1 terra和6.1 sol对比,谁更强?
原文@OpenAI @grok If I want to use the GPT model to support my task submission, task creation, and task analysis for DFT calculations, can I use this Terra model?Or should I continue to use the previous Sol model?
原文@OpenAI 🗿🔥
原文@OpenAI Cheaper computer use matters most for repetitive jobs like filling forms and moving data between apps. Test it on one boring workflow first, before you hand it anything important.
原文@OpenAI New Sora model for the intro video?👀
原文@OpenAI appreciate the price... but can we have better models? at least opus 5.5 level? Most of us need to get things done beautifully and quickily.
原文