DarkSkY

i am really human not block me

[DarkSkY @LightSkyrrr]

@uudonX 200->100

[DarkSkY @LightSkyrrr]

@ClaudeDevs win

[DarkSkY @LightSkyrrr]

Congrats to Meta on shipping Muse Wonder 🎉 A frontier model built for personal and professional agents. Huge step up for Muse.
Haven't tried Muse yet? Join at https://t.co/V7iMLCQpCg and redeem my code within 48 hours in Settings. We both get 1 billion Muse tokens 🚀
Code: C2ZMBD
#MuseWonder #Muse #AIagents

[DarkSkY @LightSkyrrr]

@leo114119 然后就降智

[DarkSkY @LightSkyrrr]

@leo114119 只靠前端效果评估并不客观

[DarkSkY @LightSkyrrr]

@MagicPower21M 有没有可能决策不是他做的

[DarkSkY @LightSkyrrr]

@thsottiaux If the quality and usage of the service can be guaranteed, I don't mind introducing a 5-hour time limit. But repeated changes have brought about credit problems. We Need A Promise Tibo

[DarkSkY @LightSkyrrr]

@bourneliu66 next level

[DarkSkY @LightSkyrrr]

@DLKFZWilliam2 They really need to focus on bringing the output quality back up. The model has been nerfed so badly lately. At its core, this is a compute shortage issue—you just can't have high usage, top quality, and no time limits all at once.

[DarkSkY @LightSkyrrr]

@Ananth7e OpenAI’s 'inflated' claims are just unacceptable. Honestly, having a 5-hour limit might actually make compute allocation more reasonable and improve output quality. If that’s the case, I’m all for adding the limit

[DarkSkY @LightSkyrrr]

@thsottiaux Honestly.If we can get Claude-level quality, I'd much prefer a 5-hour limit that doubles the usage allowance. Claude has never dropped its 5-hour window, making OpenAI’s 'inflated' claims completely unacceptable.

[DarkSkY @LightSkyrrr]

@xodud_rkd If we can get Claude-level quality, I'd much prefer a 5-hour limit that doubles the usage allowance. Claude has never dropped its 5-hour window, making OpenAI’s 'inflated' claims completely unacceptable.

[DarkSkY @LightSkyrrr]

@KRVR_AVP not bad

[DarkSkY @LightSkyrrr]

@thsottiaux we need both, love U Tibo

[DarkSkY @LightSkyrrr]

@DailyHoukaIzumi me toooo

[DarkSkY @LightSkyrrr]

@excwr 这卡池的概率100%有问题，根本没有up概率提升

[DarkSkY @LightSkyrrr]

@siratama_z 确实啊，根本就不存在up，我居然在这个池子抽到两个yuno，卡池角色越来越多，就越难抽到当期角色。最糟糕卡池机制

[DarkSkY @LightSkyrrr]

@thsottiaux codex is best

[DarkSkY @LightSkyrrr]

@AnthropicAI OpenAI is done

[DarkSkY @LightSkyrrr]

@luoling8192 最垃圾的UI模型