John Markwell

NLP and CV fans

[John Markwell @markwell_john]

@AndrewYNg @Mononofu @JensenHuang Exactly

[Andrew Ng @AndrewYNg]

@Mononofu @JensenHuang This is a false equivalence. Everyone has the right to keep their code private. The problem is when someone tries to stop OTHERS from open sourcing.

[sarah guo @saranormous]

Caught up with @karpathy for a new @NoPriorsPod: on the phase shift in engineering, AI psychosis, claws, AutoResearch, the opportunity for a SETI-at-Home like movement in AI, the model landscape, and second order effects

02:55 - What Capability Limits Remain?
06:15 - What Mastery of Coding Agents Looks Like
11:16 - Second Order Effects of Coding Agents
15:51 - Why AutoResearch
22:45 - Relevant Skills in the AI Era
28:25 - Model Speciation
32:30 - Collaboration Surfaces for Humans and AI
37:28 - Analysis of Jobs Market Data
48:25 - Open vs. Closed Source Models
53:51 - Autonomous Robotics and Atoms
1:00:59 - MicroGPT and Agentic Education
1:05:40 - End Thoughts

[Kimi.ai @Kimi_Moonshot]

Zhilin's full GTC 2026 keynote is here.

If you're curious about the "how" behind scaling Kimi’s latest models, this is the session you can't miss. :)

[John Markwell @markwell_john]

@Bandol8848 @XPengMotors Target is a slim chance.

[John Markwell @markwell_john]

@xiaopenghexpeng Don't be late, VLA 2.0.

[沉浸式翻译 @immersivetran]

悄然间，我们已经一起走到了3万粉丝这个小小里程碑。

说句掏心窝子的话，是因为你们的使用、反馈和每一次互动，这个数字才有意义。

多的不说了，关注+转发+点赞这条帖子

随机抽 30 位小伙伴

赠送价值 207 元的浸式翻译pro版季卡（3个��）

🫶

[沉浸式翻译 @immersivetran]

最近在 Reddit 上看到一篇很难忽略的帖子，一位自称在 FAANG 和独角兽工作过的工程师，记录了自己对 AI 进展的心理转变。不是恐慌式的“要被取代了”，而是一种更冷静、也更令人不安的接受。

他描述的现象很具体：和 Claude 配合完成原本需要数周的功能，已经变成日常；“vibe coding”从几个月前的段子，变成生产环境里的现实操作；人类在工作中的角色，越来越像是给模型补上下文、做 sanity check。正如 Karpathy 说的那样，“程序员真正贡献的 bits，正在变得越来越稀疏。”

真正让他崩塌的并不只是写代码。一次就诊前，他把症状输入模型，生成了完整的用药选项和利弊分析；在诊室里，他用模型给出的理由反驳了医生的第一建议，对方很快改变了方案。他第一次认真地问自己：以后和“专家”对话，是否还不如直接和模型对话？

问题开始浮出水面：如果模型继续进步，解决了上下文窗口、持续学习（Continual Learning）这些最后的限制，人类当前在知识工作中提供的大部分价值——理解背景、整合信息、形成判断——是否都会被系统性地压缩掉？而这并不是边缘人的幻想，Anthropic 和 DeepMind 的研究员们正在公开表达类似的困惑：有人转向“自动���对齐研究”，有人坦言“不知道接下来该在乎什么”。

评论区的反应同样耐人寻味。有人说自己还保有工作，但心理上已经转向“稳定优先于优化”，不再把学历、里程碑当成可靠的未来信号；也有人用黑洞事件视界作比喻：我们正坐在一艘飞船上，越过不可逆的边界，不知道另一侧是毁灭，还是完全超出想象的世界——但这可能是人类历史上最重要的一次第一人称体验。

这篇帖子最刺痛人的一句话或许是：真正的问题不是“你还能做什么”，而是“人类参与是否正在让系统变差”。

原帖⬇️：（建议结合评论一起看）
https://t.co/BAmz3dbR4F

[John Markwell @markwell_john]

XiaoPeng should buy a social media like weibo.#xpev

[沉浸式翻译 @immersivetran]

不知不觉2万粉了。
说再多感谢也不如发一波福利
关注+转发+点赞这条帖子
随机抽20位小伙伴
赠送价值207元的浸式翻译pro版季卡（3个月）

[Milvus @milvusio]

🚀 Excited to announce a tutorial on building a RAG pipeline with Milvus and @SiliconFlowAI 🌐

🔍 SiliconFlow's scalable AI infrastructure integrates with Milvus' vector database to enhance your data processing capabilities.
🧠 Learn how to deploy LLMs and embedding models effortlessly with SiliconCloud's MaaS platform.

📚 Use our step-by-step guide and start building your RAG pipeline today! 👉 https://t.co/5KlPtfwjAZ

#Milvus #SiliconFlow #AI #RAG

[John Markwell @markwell_john]

@_akhaliq Great

[John Markwell @markwell_john]

@JinheonBaek @MSFTResearch Cons😁

[Barsee 🐶 @heyBarsee]

1. https://t.co/Nei3IxXotQ by @lovolabs

Text-to-speech that creates realistic voiceovers for videos, podcasts, audiobooks, and more.

Listen to this voiceover 👇

[elvis @omarsar0]

One of the main uses of ChatGPT I've seen is information extraction.

This paper provides a deeper assessment of ChatGPT's performance on this important task.

https://t.co/MYaqIc1NMz

[Jeremy Howard @jeremyphoward]

There's a lot of folks under the misunderstanding that it's now possible to run a 30B param LLM in <6GB, based on this GitHub discussion.

This is not the case. Understanding why gives us a chance to learn a lot of interesting stuff! 🧵
https://t.co/9xOfNyLUFr

[Yi Tay @YiTayML]

Amazing work Sebastian!

[Daniel van Strien @vanstriendaniel]

⚖️ On average, datasets on the @huggingface hub are used to train 0.94 models hosted on the hub
⬆️ The maximum number of models trained on a dataset on the hub is 1,332 🤯
🔝The top 20 used datasets may be familiar to you!

[Andrew White 🐦‍⬛ @andrewwhite01]

New features in paper Q&A @Gradio space:
1. You can now mix repos and papers, so you can ask questions about a research paper by having the code implementation and paper combined
2. Cost cut by 10x thanks to ChatGPT integration
3. Performance seems to be similar or better 1/2

[John Markwell @markwell_john]

@stefan_it_ @huggingface @AmazonScience Great