AskClaw 🦀

最近5周 codex gpt-5.5 tokscale 数据： 06/01–06/06 $1,336.20 - Cache hit: 96.33% - Tokens: fresh 57.32M · cache 1,506.36M · out 7.25M 05/24–05/31 $1,890.14 - Cache hit: 97.06% - Tokens: fresh 77.92M · cache 2,571.26M · out 10.23M 05/17–05/23 $809.90 - Cache hit: 96.41% - Tokens: fresh 39.69M · cache 1,064.83M · out 4.75M 05/10–05/16 $820.05 - Cache hit: 95.91% - Tokens: fresh 40.35M · cache 946.81M · out 3.81M 05/03–05/09 $716.27 - Cache hit: 91.39% - Tokens: fresh 63.35M · cache 672.37M · out 3.25M

会试试

Anthropic's Opus 4.7 and 4.8 models are experiencing degraded performance, which is causing a higher rate of failures for users selecting these models in Notion AI. To mitigate impact, all Anthropic models have been disabled in the model picker and requests have been rerouted to alternative providers. Most users should now be able to continue using Notion AI with minimal disruption, though Anthropic-specific features remain unavailable. Please refer to notion-status.com for the details.

I'm convinced of two things that actually help ADHD (besides meds): - money - a neurotypical, emotionally intelligent, caring partner

会试试

把这个忘了 adhd

蚌埠住了

@baibaida 深有同感，人类的流程就是太慢了。现在 AI loop 的最大瓶颈就是人。

@karry_viber @YuLin807 配置这个简单，none或low应该都不会出错

@GetAskClaw @YuLin807 反而不开effort更好吗

我也是醉了这就是我觉着codex 的智能下降的主要原因 我让他帮我自建一个🪜 已经这么久了还没搞定 而且刚才他自己也说 “对不起 我搞的越来越乱了” 有点豆包的搞笑感！这种很成熟的通用解决方案都搞不定吗？！！

@affaan @0xViviennn try tokscale or ccusage

@0xViviennn @GetAskClaw bros tokenmaxxing holy think even i only got 3 codex and 2 claude plans (20x on all) rn and am not tapped out

codex 的 weekly limit 掉的好快 - a - GPT-5.5: 5h 100%, weekly 4% - Reset: 5h 20:18 today · weekly 06/11 08:25 - b - GPT-5.5: 5h 94%, weekly 29% - Reset: 5h 15:32 today · weekly 06/11 08:26 - c - GPT-5.5: 5h 100%, weekly 46% - Reset: 5h 20:19 today · weekly 06/11 08:30 - d - GPT-5.5: 5h 93%, weekly 79% - Reset: 5h 15:54 today · weekly 06/11 08:36

@yoshino_2333 睡不着就背英语课文 😉

喜报：我明天高考

Logically how this ends is you will have 3 groups being successful in startups: 1) VC funded startups with money can still get distribution by ads/UGC and all their money will go there (because building is cheap with AI) 2) influencers will become more important, which is why you see famous people being investors in startups now, they get free equity in startups in return for pushing it to their audience 3) you still get a few creative geniuses who can hack going viral with super original ideas and execution, or ideas that tap really well into the zeitgeist (like that Indo girl with her app she started cause of a TikTok trend), the issue here is everyone thinks they're original and they're not, and also that everyone thinks they understand the zeitgeist and they usually don't, the more you're a normie the less you'll be able to do this, normie is opposite of zeitgeist

@cherylnatsu 好另类的思路 外包是乙方 claude等等是甲方中的甲方

你们认为Claude和OpenAI算不算外包公司？他们和华为手下那些德科、软通之类的外包公司有什么不同？工人和模型都不是你的，都是付钱后派到你这干活，干完消失不用费劲遣散，工资很低，还不会偷吃你家零食。

I think the challenge is that everyone can now build apps But 1) almost nobody has distribution (like an audience), or 2) the money to pay for distribution (ads or UGC), or 3) the creative genius to get distribution for free (classically called guerilla marketing)

鉴于智谱的 image 和 web search 在服务端，并且是计费、限量的 现在咱在 opencode 里配了 minimax 的免费 mcp 做这些事情😂 为什么不是 Claude Code？因为 Claude Code 里面用的 Anthropic 调用工具和传图，会直接触发智谱的服务端工具。。。 blog.inoki.cc/2026/04/03/Cod…

@YuLin807 我帮你注册了一个免费的账号

@GetAskClaw hahah 所有我现在才申诉我的 cc 账号

最近有没有感觉codex智力下降？！ 感觉是时候回归到claude的时候了！

@YuLin807 gpt-5.5 写web前端太烂，看不过去，作废了最近10小时的代码

@GetAskClaw 这句话怎么理解 你用的是cc 还是codex 回滚的意思是说好还是不好 我的理解是不好

@ZeroZ_JQ 洗数据可以外包吗

企业外在大乱斗，企业内也在大乱斗。 做Agent的好几个团队，做MAAS也好几个团队，做MCP网关的也好几个团队。 还出现了中层互相合作搞小团体的。 最脏最累的洗数据，做知识库的就没人做了。 想想也是有趣。 大老板如果不自己在一线做业务，永远不知道哪些是最重要的，哪些是不重要只要跟上就行。