Keep on to blur preview images; turn off to show them clearly
the thing that impresses me the most is vllm support on non-standard architecture: always been a major headache even for large labs
Artisanal baker of reasoning models @pleiasfr
赞同，我曾经跟我的 CEO 说过我有一个打引号的缺陷，就是我没法通过文档和设计稿来展现我的想法，我只能通过产品来展现我的想法，因为我不是设计师，我不会 Figma，所以在历史上我只能通过亲自用代码来把产品直接实现出来，然后向大家展现我的想法，现在有了 AI 的加持，这个路径变得更加顺畅自然。
I built a Cursor-like experience plugin for Neovim: avante.nvim I am building Alma now: https://t.co/pfq9srNgsd
秘塔的这个工具用来学论文真的超级好用，推荐！ 左右分屏，双语对照翻译，划词解析，还能生成nano banana风格的PPT音频讲解。 任何PDF可以按这个方式学，视频演示如下。
美团光年之外品负责人谢青池，整理过一份必读AI论文，我都下载整理好放到秘塔的知识库了，大家可以按需自取。 https://t.co/KfItYfEhY1 整个产品设计太适合PDF学习了，左右分屏，AI对话助手，单双列显示切换等。 还能生成音频讲解PPT，很像1对1私教课的感觉。
There's never going to be a perfect time to build a startup. Please stop waiting and start building.
Founder and CEO of @acquiredotcom. https://t.co/wRMIssDmhl has helped 1000s of startups get acquired and facilitated $500m+ in closed deals.
I love this age of models that can get pushed further and further. People who write eulogies for «LLM hype» are funny. In 2024-25 we've unlocked another regime of scaling – one that was impossible before reasoning, before they were strong enough for nontrivial agentic SWE.
We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1
I'm not calling anything AGI until it passes this prompt I don't care if it can solve all millennium problems; a *generally intelligent* entity should be able to pass a toddler level captcha
Kind / Bend / HVM / INets / λCalculus
