Two weeks without mobile internet improved mental health more than antidepressants and reversed roughly 10 years of attentional decline. Screen time dropped 49% (314 to 161 min/day). https://t.co/9sU6ZFXDCr

The single greatest thing I did for my screen time was get a @thetinypod. It's an iPod case for your Apple Watch. Take calls, listen to music, get GPS directions The only thing you can't do is mindlessly scroll the internet. Actually genius https://t.co/1kvEUxqc77

@FreyaHolmer Dangerous knowledge 😂 Bonus UV flashlight use: When you get a chance, take it out at night in nature - plenty of lichens, flowers and critters fluoresce amazing colors! https://t.co/JIGV69Erju

Oh look! Anthropic's entire "we are delaying Mythos" narrative was marketing hogwash. Kudos to FT for confirming what was obvious. Anthropic simply doesn't have the compute. FT: "Multiple people with knowledge of the matter suggested Anthropic was holding back from a wider release until it could reliably serve the model to customers."

Anthropic’s Mythos is being positioned as a highly advanced, potentially risky model. While the company warns it may be too powerful for public release, some experts argue the concerns should be taken seriously but not overstated. The situation reflects a familiar pattern. Breakthroughs in AI often come with both real risks and amplified narratives around them. https://t.co/7xmADAltal @stokel @sciam

@genesobolev My wife is creating educational songs for kids. She puts a lot of love into it and I think it comes out well Of course it takes lots of effort to specify carefully in natural language https://t.co/KLIg1Z86wf

With my infinite agent canvas, I can spin up more designs that build on the others. You can imagine just generating and editing all kinds of artifacts with a similar theme in a matter of minutes. Not the best example below, but I think it communicates the vision here. https://t.co/pa1yiRDgsR

STARLINK: SpaceX Starlink Government & Enterprise for Latin America and the Caribbean, Juan David Vélez (@JuanDaVelez); and Starlink Sr. Director of Market Development, Jonathan Kleinman (@jonathankleinma), presented a comprehensive connectivity proposal to Honduras President Nasry Asfura. The Honduran government is promoting an alliance with SpaceX to connect 8,000 schools before the end of the 2026 school year! More Starlink news in the ELON CHRON below!

I prefer my design tool to be more closely integrated with where my agents work. I spent a few hours building my own design tool (inspired by Claude Design) inside my orchestrator. I can use this with Codex, Claude, or any model. Opus 4.7 works best for now. Doing more tests. https://t.co/zcfk5KyG2L

> 1980: John Searle explains why we can't abstract away the causal properties that actually produce mind > 2025: Minds, Brains, and "but what if we scaled the program" > 2026: Twitter still thinks simulated water is wet when argument is rehashed > 2035: Sam Altman: "ok fine it was autocomplete the whole time" > 2045: Chalmers: "the hard problem was, in fact, hard" > 2050: textbooks: "the 2020s functionalism revival is now considered an embarrassing episode, like phrenology"

Platforms are starting to verify something we once took for granted, being human. Apps like Tinder and Zoom are introducing eye-scanning “proof of humanity” as AI-generated identities become harder to detect. It marks a shift. In the AI era, trust may begin with proving you are real. https://t.co/46P5JcnWPp @bbcnews

What's your most used comment, and why is it LGTM? Get your t-shirt over at the GitHub Shop. https://t.co/IeViT573aZ https://t.co/zLJYzUn6Dw

Gave codex my plasmid design notebook and asked it to animate the two key proteins doing their thing, for a video explanation I'm working on. Pretty much nailed it! Fun to wake up to work done :) https://t.co/nmE8D0ZXxj

Love this one even more. All done with Hermes Agent. cc: @Teknium @NousResearch https://t.co/Os46eK7lV7

Love this one even more. All done with Hermes Agent. cc: @Teknium @NousResearch https://t.co/Os46eK7lV7

AI is entering newsrooms, but not without resistance. While it promises efficiency, critics argue it risks diluting voice, judgment and the core craft of storytelling. The trade-off is deeper than it looks. Speed can come at the cost of meaning. https://t.co/4baYjjkImS @wired @StevenLevy

@SchrodingrsBrat some people are saying https://t.co/Ivjm9rrnI6

在 Ollama 运行 hermes https://t.co/yU4ywysy5Z

在 Ollama 运行 hermes https://t.co/yU4ywysy5Z

Hermes Desktop v0.5.0 – native Mac app for Hermes Agent. SSH only. multi-profile + full cron manager, now live ✅ NO gateway, NO daemon, remote is source of truth. The official dashboard just shipped too – great for browser admin! @Teknium @NousResearch https://t.co/19mgR9Aoi7

Trying to write a tutorial for hermes, anyone interested? https://t.co/eZDnEgtZGy

Trying to write a tutorial for hermes, anyone interested? https://t.co/eZDnEgtZGy

@trendingbots @SkylerMiao7 @MiniMax_AI https://t.co/zJpv3eIMWC

🚀 Excited to share ViPRA: Video Prediction for Robot Actions 📍 Accepted to #ICLR2026 @iclr_conf 🏆 Best Paper — #NeurIPS2025 Embodied World Models Workshop Robot learning today still needs millions of action labeled videos. Yet videos are abundant — from humans and the web — but lack action labels. Meanwhile, pretrained video models already learn rich dynamics. ViPRA is a recipe for turning pretrained video models into robot policies while enabling robot learning to scale with actionless videos. 🧵 Thread ↓

@0xViviennn Merged! https://t.co/h9grjwm8ua

在用 Hermes Agent 的时候，一直感觉跨 session 的中文搜索不太好用。Agent明明说过的话，session_search 就是搜不到，像完全失忆了一样.... 比如昨天搞了一天的A2A通信，但今天再搜"和其他Agent的聊天记录"的时候，返回 0 条结果。但数据确实在数据库里——用 LIKE子串匹配能搜出 20 多条 然后我让 Claude Code 帮忙查了一下，原因是 SQLite FTS5的默认分词器。英文有空格，天然能按词分 中文没有空格，FTS5把每个汉字当成独立 token。搜"记忆断裂"实际执行的是 记 & 忆 & 断 & 裂——四个独立的字分别匹配，丢失了词的连续性，搜不到原本存在的内容 日文、韩文也一样，都没有空格分隔，都会中招，但英文用户完全不受影响。 目前的解决方案是 FTS5 搜不到结果时，检测到查询包含 CJK 字符就自动降级到 LIKE子串匹配。比 FTS5 慢一点（全表扫描），但对 agent 的数据量完全够用。 已提 PR： https://t.co/QgT9roKg6P @NousResearch @Teknium

Thanks to our incredibly community and Viviennn in particular, we have greatly improved session-based recall for Chinese users! Thanks for the contribution! https://t.co/h9grjwm8ua

Hermes 一丢 Agent，全网又卷出 5 个新进化体！ Nous Research 的 hermes-agent（96k+ stars）底层太能打了：持久记忆 + 自动提炼技能 + 跨会话成长，社区直接当 DNA 疯狂 remix。 Atlas 上已经 90+ 项目，这次我挑了5 个最新最炸的进化体（全避开老面孔），AI 玩家看了会沉默，Agent 爱好者看了会原地 fork： 1️⃣ hermes-webui（https://t.co/EYpeGa1g4w） 浏览器/Web + 手机端 UI，把 Hermes 进化成“随时随地指挥”的 Web 版。暗黑风响应式，X 上直呼“手机党终于翻身了！” 2️⃣ hermes-dashboard（https://t.co/Y3insgftNF） 实时 Web 监控台 + 自动 Wiki：多 Agent 并行、工具调用、灵魂状态一屏看尽，还把 memory 自动转知识库。生产环境标配，Atlas 安全审查通过！ 3️⃣ hermesclaw（https://t.co/WJzA56oSjv） 官方主 repo 重点推的 WeChat 桥接器，让 Hermes 和 OpenClaw 共用一个微信号，双向无缝。中文玩家狂喜：“微信生态直接被 Hermes 入侵！” 4️⃣ hermes-hudui（https://t.co/V3AZcUQCVK） 原 TUI HUD 的 Web 进化版，浏览器里实时看 Agent “在想啥”、持久记忆流动。灵魂监控 2.0，推特名场面：“终于看到 Agent 的内心戏了”。 5️⃣ awesome-hermes-agent（https://t.co/2QVFSb4Dlb） 社区维护的“进化树索引”：技能、插件、集成、教程全收录（已近 900 stars）。Atlas 列为 Guides 头牌，想 fork 就从这里抄作业。 🟢 为什么这些新项目这么爆？ Hermes 从不给你黑箱，它给的是可 hack、可扩展、可自我改进的底层循环。你 fork 它，不是在用工具，而是在和 Agent 一起递归长大。 2026 年开源 Agent 的正确姿势，就是把 Hermes 当骨架，自己长肉！

+3 more

Anthropic’s Claude Mythos is drawing attention for its advanced capabilities. The model is reported to outperform humans in certain cybersecurity and hacking-related tasks, raising concerns among regulators and institutions about potential misuse. The risk is clear. As AI becomes more capable, the line between defense and offense becomes harder to control. https://t.co/v7tig74Aw9 @bbcnews @liv_mcmahon @joetidy @bbctech

The game has really changed in animation. This is my latest project Kyōdai (きょうだい ) AKA → "Siblings." An AI film concept I'm developing @promise_ai While not perfect by any means, tools like Seedance 2.0 allow for rapid prototyping and visual look development. Once explored, we bring in artists to hand draw and create the world to build and protect the IP. The VFX artist to build out the world with a hybrid 3D and AI methodology. This is the world we live in. What a time.

@LinusEkenstam Yes and https://t.co/kiuZ7QXLzb has all the good stuff from X

TBPN asked me how Zapier keeps AI from drowning in data. Short answer: we built it a brain. Three layers: 1. Company-level source of truth (strategy, values, ICP). Curated by me and a handful of senior leaders 2. Team-level context that cascades down 3. Individual context, private to each person. Meeting transcripts, Slack threads, project docs, etc. So when anyone at @Zapier talks to AI, they're not starting from scratch. They point the SDK at specific documents against the backdrop of that brain. Better inputs, better AI. Thanks to @TBPN for having me on