Your curated collection of saved posts and media
After the Reform Party HUMILIATED Labour in elections, Keir Starmer smeared the ‘Unite the Kingdom’ rally as “extremist” and banned Conservatives from entering the country. The people of England’s response?: Tens of thousands FLOODING the streets 🇬🇧🏴 https://t.co/lNTxhLeT7q
Everyone has seen the @waitbutwhy cartoon of AI capability growth with a "you are here" indicator just before the exponential really starts, but the independent assessments of both METR and the UK's AISA do seem to show that we are past that point now (until we hit a slowdown?) https://t.co/vxYc6GSS5d

NVIDIA just released a paper review dataset on Hugging Face APRES, Agents4Science, and Sakana v2 subsets covering human and AI-authored papers with real review decisions. https://t.co/louiNyRGC5
AI will make translation invisible. But language was never just about translating words. Language is trust. Culture. Humor. Negotiation. Influence. The people who understand this will still have a major advantage in business and life. AI removes friction. Human connection remains the moat. https://t.co/KjuTAKxG6x @ConversationEDU @ConversationUS
New article: a visual tour of recent LLM architecture advances, from Gemma 4 to DeepSeek V4. I focus on long-context efficiency tweaks like KV sharing, per-layer embeddings, layer-wise attention budgets, compressed attention, and mHC. Link: https://t.co/KO81y3kTH7 https://t.co/wTx51QpQu4
Your Grok Subscriptions work inside of Hermes! https://t.co/TGbJMDzTMI
Your Grok Subscriptions work inside of Hermes! https://t.co/TGbJMDzTMI
GPTとseedance 2.0で全自動ショートアニメ 気になって初めて試してみたけど確かに企画だけすればGPTがキャラとストーリーとコンテ考えてくれて一切何もしなくても全自動でアニメっぽくはなるね 現状コンテが酷いのでもう少し教える必要がありそうだけどこんなに簡単だと色々変わる気がする 次はもっと精度上げてみようかな
tesla robotaxis going about as well you might expect. https://t.co/0Ii8BdsLJQ
北海道出身! ClariSのクララさんがファーストピッチに挑戦⚾ #lovefighters @ClariS_Staff https://t.co/7GwM20arnM
北海道出身! ClariSのクララさんがファーストピッチに挑戦⚾ #lovefighters @ClariS_Staff https://t.co/7GwM20arnM
This is amazing - we have Singapore's Minister of Foreign Affairs @VivianBala explaining how he uses a Nanoclaw on Raspberry PI This is a gem; He says the barriers to accessibility have collapsed - his setup was not created by him; and that memory is the next frontier (LOL) https://t.co/ZKhV8ExGx5

LLM-powered AI agents are gonna be great! You should totally trust them! https://t.co/J8d6kZdinI
Anthropic is through almost $10 billion a month and thus needs more funding before it will do its planned IPO this year. It and its rival OpenAI have collectively raised more than $220 billion in funding since beginning of last year, which tells us a lot about their cash burn https://t.co/8VDqGOJM0E
we made tech fun again 🫡 s/o to @adilmania @siliconmania for the primo merch https://t.co/gdWPEAWWXw
from SF to Singapore 🫶 make tech fun again.
we made tech fun again 🫡 s/o to @adilmania @siliconmania for the primo merch https://t.co/gdWPEAWWXw
Hermes Web UI v0.5.25 发布。 群聊房间支持重置 / 克隆 Web UI 状态目录可配置 新增 MiMo TTS (普通话/台湾话 超好听/粤语/英语) Provider 模型列表改由后端代理,减少跨域失败 修复 Bridge 工具授权 base64 图片历史可正常展示 补强 Playwright、流式契约和模型测试
Hermes 又集体进化了! 技能超市、企业员工适配器、本地记忆黑科技、创意超级英雄工作流、Obsidian 身份层……全网程序员把 Hermes 玩成了下一代 Agent 商场 + 公司员工 + 懂你大脑 + 创意工厂 + 私人知识永存神器: 1️⃣ hermeshub(https://t.co/CQH8Kpmpeo) 社区技能注册表 + 安全扫描市场,一键 install + x402 付费。 “Agent 技能也能逛淘宝”实锤了! 2️⃣ hermes-paperclip-adapter(https://t.co/oBRHtQn2Kt) Paperclip 官方适配器,让 Hermes 变身企业“受管员工”。 任务分发、技能同步全自动,企业党狂喜! 3️⃣ hermes-agent-yantrikdb-memory-plugin(https://t.co/hMIMlisRFX) 本地动态 memory 插件,skill 生成 + recall 零 token 浪费。 记忆从“存”进化成“真懂你”😂 4️⃣ superpowers(https://t.co/WsqeCQRoJV) 创意工作流包,设计哲学 + 品牌 pipeline 打包成 skill。 Agent 也要当创意超级英雄了! 5️⃣ obsidian-skills(https://t.co/CUtucuUU6P) 30 秒把 Obsidian vault 变成 Agent 身份层。 私人笔记直接变“自我认知”上下文,隐私党福音!
Hermes 继续自我完善,冲最强 Agent! 强化 fork 版、23 插件管家、一句话组件工厂、像素世界桥接、WeChat 多 Agent 桥接…… 全网程序员把 Hermes 玩成了下一代 Agent 深度 hack 神器 + 管家大脑 + 组件工厂 + 元宇宙居民 + 微信军团: 1️⃣ hermes-agent-fork(https://t.co/YxXA60KBDx) 强化 fork 版,保留自进化循环 + OpenClaw 迁移优化 + 子 Agent 并行。 深度玩家直接上手 hack! 2️⃣ hermes-plugins(h

Bye bye claw? At least for now, all ThursdAI cohosts have moved to Hermes / Codex https://t.co/cTPm35xsah
Introducing v0 Browser Use. v0 can now open the apps it builds, use them, critique designs, debug complex flows, and fix things proactively. While it's working, v0 will send you screenshots of what it sees. https://t.co/lPdFSbZSDq
【保存版】 「Hermes Agent で会社1個まるごと、VPS 1台で動かす」事例が話題👀 この自社の組織図、4層に分かれてこんな構成になっている👇 1. Company Brain → ビジョン / ブランド / 顧客 / プロダクトの永続コンテキスト → 他の全層がここから情報を継承 2. Orchestrator Hermes Agent → Company Brain を読んで、適切な部署エージェントへ振り分け 3. 各部署エージェント(Marketing / Sales / Eng / Ops 等) → 役割別に独立コンテナ、互いに干渉しない 4. ツール層(Web fetch / DB / Slack / Email) → 共通インフラとして全部署が呼ぶ つまり 「人間1人 + Hermes Agent 4層で 中規模スタートアップの内部構造を再現」 できてる、という事例。 OpenClaw でメモリ層 / ナレッジ層を真面目に運用してる人なら この組織化の発想、そのまま転用できる。 週末で1個試す価値ある構成。
@bwjbuild @Jaaneek Who said you could? You can as of 20 minutes ago but you couldn't when we launched this... It will auto enable if you use supergrok oauth or put an x api key into the .env - use `hermes tools` on latest hermes https://t.co/I8B76FB99a
@abboskhonovv @NousResearch Nice be sure to share in the community projects forum in our discord for more visibility! https://t.co/5EoJ4EBecb
You can now use your ChatGPT subscription in the Zed agent, with the same usage and rate limits you benefit from in Codex directly. We're grateful that @openaidevs continues to support subscription-based access for third-party tools, even as others move toward usage-based billing.
Fei-Fei Li warns that AI may be staring too hard at language models. The world is not just text on a screen. It is physical, visual, spatial, and always changing. Most of the economy runs on seeing, moving, interacting, and embodied intelligence. https://t.co/mGMjsCdU34
🛑[LIVE] on day 2 of @aiDotEngineer 🇸🇬 @ZixuanLi_ from @Zai_org on handling long-horizon tasks requiring adaptive planning and engineering insights. “You have to learn from your previous failures” https://t.co/jsO7FwSdmu

🛑[LIVE] on day 2 of @aiDotEngineer 🇸🇬 Daria Soboleva from @cerebras on building mixture of experts at scale. “GPU scaling hits a wall when communication becomes the bottleneck. The real breakthrough is keeping massive expert networks on-chip without model partitioning.” https://t.co/HYZpBeWtIZ

🛑[LIVE] on day 2 of @aiDotEngineer 🇸🇬 Andrew Tan from @GroqInc on scaling low-latency LLM inference at GroqCloud. “Enterprises are increasingly looking for dedicated compute capacity.” https://t.co/lV8moysjJv

🛑[LIVE] on day 2 of @aiDotEngineer 🇸🇬 @JanLiphardt from @openmind_agi on building the android moment for robots. “Our eval criteria - smiles and tears and trust and memories” https://t.co/xftKn4pZ4c

🛑[LIVE] on day 2 of @aiDotEngineer 🇸🇬 @taiuti from @reactorworld on how AI is moving from generating content to simulating reality itself. It’s unlocking new interfaces for robotics, creativity, entertainment, and real-time interaction. https://t.co/nKFedO00Tw

🛑[LIVE] on day 2 of @aiDotEngineer 🇸🇬 Selim Arguel, robotics engineer at @menloresearch on engineering sim-to-real for open source humanoids “Noise is all you need” https://t.co/318m81fPI5

🛑[LIVE] on day 2 of @aiDotEngineer 🇸🇬 @ryolu_, head of design at @Cursor on a future where software is collaborative, customised and shared amongst humans. “The future belongs to people who build to think” https://t.co/F3eAV0VTTe
