personal-agentPOST
用户吐槽 OpenAI Dots 执行任务不彻底
开发者称 Dots 是「低投入模型」,收件箱例行任务只做一半,后台过程不可见,认为 Grok Bot 或更靠谱、Muse 体验最好但同样受限。
David Cramer
@zeeg20 天前收录原推发布于 2026.10.17
David Cramer @zeeg原推发布于 2026.10.17
Biggest annoyance with OpenAI’s Dots: it’s a low effort model. It reminds me of a year ago when using models was non stop prompting for the next step. I cant get it to reliable achieve tasks due to this. For example my inbox routine it will just do a bit of the work and completely ignore the rest. On top of that because all background work is hidden (afaict) as a user I don’t even know if it’s doing anything for what it’s doing. I know Grok Bot will use better models in some scenarios so it might be the best solution right now. Muse imo is the best UX but it has some of the same limitations as Dots.
x.com/zeeg/status/2111131126924550508采集于 2026.10.17 10:29