Podlens_Jason

Building PodLens | AI podcast search for humans and agents.

[Thariq @trq212]

We removed ~80% of the Claude Code system prompt for our newest models, this is what we've learned about writing system prompts, skills and Claude.MDs for them. https://t.co/6DZwSrZjE9

[Podlens_Jason @Delibread2]

@realWeZZard 从发言的内容来看，梁圣不如杨植麟，AI 公司 CEO 的技术含金量太重要了，看好 Kimi 和 Anthropic。

[Podlens_Jason @Delibread2]

@arkuy99 Opus 5要发了。

[Podlens_Jason @Delibread2]

@ezshine 都1000万活跃用户了，这个时候还不降智说不过去🤣

[Andrej Karpathy @karpathy]

One pattern I find useful for working with LLMs is a nice long ramble session. Sometimes the LLM needs more bits to understand what you're trying to achieve, but you're too lazy to type them. In these cases I like to lean back, switch to /voice and just ramble for like 10 minutes, total mess, anything goes, full stream of consciousness. Sometimes I declare it up top, something like "switching to speech recognition sorry for any typos...". Sometimes I turn it into a small interview of a few turns. But I find that the LLMs are somehow very good at reconstructing long incoherent rambles and often their echo of your own tangle of thoughts comes out quite a bit cleaner than what you started with. The result is that you improve the mind meld and have to correct things less from that point on.

[Podlens_Jason @Delibread2]

Trending！

[unusual_whales @unusual_whales]

BREAKING: Trump is overhauling federal science funding by directing the roughly $200 billion annual research budget away from traditional university grants and toward artificial intelligence and individual scientists, per WSJ

[unusual_whales @unusual_whales]

BREAKING: Trump is overhauling federal science funding by directing the roughly $200 billion annual research budget away from traditional university grants and toward artificial intelligence and individual scientists, per WSJ

[Podlens_Jason @Delibread2]

@haruhikoJoe 真小人与伪君子的区别。真小人>伪君子！

[Podlens_Jason @Delibread2]

AI turns scarce expert exploration into massive, parallel, verifiable search. It won't solve everything—but it may reveal which problems need genuinely new ideas, and which answers were simply hiding in corners we never searched.

[Podlens_Jason @Delibread2]

@wquguru Fable 5修复后你让GPT 5.6 review也大概率一样的结果， Fable 5的后训练没有做好，执行验证不仔细，claude模型的通病。

[Datacurve @datacurve]

Kimi K3 debuts at #3 on DeepSWE.

It's the first open-weights model that delivers frontier-level performance, achieving results similar to Claude Fable and GPT-5.6 Sol.

[Podlens_Jason @Delibread2]

@DLKFZWilliam2

[Podlens_Jason @Delibread2]

@bindureddy Zero.

[Podlens_Jason @Delibread2]

@MaxForAI 不要对5.6有过度的期待，大概率不如Fable 5🤣

[Podlens_Jason @Delibread2]

@xingpt Fable 5 low> Opus 4.8 Max

[Podlens_Jason @Delibread2]

@0xLogicrw 两个原因 Musk 已经很难赶上 AO 两家了：1. xAI 已经完全丧失最顶级人才的吸引力, 2. 另外两家已经建立模型自进化 pipeline。

[Podlens_Jason @Delibread2]

@Shina_Kaede 算力都给企业用户了，个人用户是纯补贴，都没有动力了。

[Akshay 🚀 @akshay_pachaar]

https://t.co/VoQByJUrBU

[Thariq @trq212]

https://t.co/R6exTuF7P8

[Podlens_Jason @Delibread2]

@AISuperDomain Claude 意图对齐能力是独一档的存在，