Your curated collection of saved posts and media
Wow how @aiDotEngineer has grown!! More than double the size of last year. Excited to emcee here tomorrow and introduce some of my friends cc @dexhorthy https://t.co/B6cJ2sM2Fn
passed by this ad where Chet hanks is the face of a morning after pill lol https://t.co/ti8ABiRStP
Clodo website is really good https://t.co/jgpoX1WiZf
@JPoehnelt Nice I'm here as well. Swing by Thursday if you can! https://t.co/MLkvWXm955
MASSIVE NEWS Teamed up with NVIDIA to make Local AI The Default https://t.co/kmGgcBEZ4f
SITUATION REMEMBERED: 5 years ago today GitHub Copilot was released. It could complete lines of code or even entire functions, and was powered by a brand new OpenAI coding model called Codex, a fine-tuned version of GPT-3. https://t.co/eGNjqNAngZ
only OGs remember GitHub Copilot π I remember trying it out when it was released. It provided a glimpse into the agentic coding and research future we are seeing now. https://t.co/kkzckVt3ti
SITUATION REMEMBERED: 5 years ago today GitHub Copilot was released. It could complete lines of code or even entire functions, and was powered by a brand new OpenAI coding model called Codex, a fine-tuned version of GPT-3. https://t.co/eGNjqNAngZ

@thesphyramid https://t.co/xR5d9VwHB6
Karpathy calls it βAI psychosis,β but thatβs the wrong term. Psychosis = loss of touch with reality. What heβs describing is AI fatigue/brain fry, cognitive overload from directing agents. Survey (HBR/BCG, 1.5K workers) shows the same: fog, headaches, slower decisions. #AIFat
@itsthewealth4me https://t.co/xR5d9VwHB6
Karpathy calls it βAI psychosis,β but thatβs the wrong term. Psychosis = loss of touch with reality. What heβs describing is AI fatigue/brain fry, cognitive overload from directing agents. Survey (HBR/BCG, 1.5K workers) shows the same: fog, headaches, slower decisions. #AIFat
Musk is right that many problems can't be solved just by "throwing money" at them. But US aid to fight AIDS, child hunger & other global plagues has one of the highest ROIs in government. The data is clear: cutting this $ means millions will die. The question is how many. https://t.co/OilcCBIzxR
Join us in recognizing Delta's Rising Professors of 2026! π Congrats to the current and incoming assistant professors our community nominated as truly exceptional. These individuals were recognized for the quality of their research, teaching, mentorship, community involvement, and industry impact. These are the people that truly push the frontier, carefully mentor their students, and have incredibly strong research taste in their domain. We are honored to play a small role in recognizing them. https://t.co/1eBVEXnzbK

New blog post: βItβs Hard to Evalβ Is a Product Smell If you find it hard to verify AI output, chances are that your users will too! In other words, I often find that product design is the bottleneck In the post I embed three **interactive before/after examples** based on products I've helped with: 1. an AI data agent that answers business questions 2. a PE lessonβplan generator for Kβ12 teachers 3. a workersβ comp tool that drafts 50βpage medical reports I believe this is a significant issue in AI Engineering and upstream of evals! Link to post: https://t.co/ErA9dp4ZPw Note: I'm not a designer so the design sketches are far from perfect, but I felt it was important enough to spend a significant amount of time on this. Thanks to @sh_reya and @isaac_flath for feedback.
One day left to submit your projects for the Hermes Agent Accelerated Business Hackathon presented by @NVIDIAAI Γ @stripe Γ @NousResearch! Submissions close 11:59 PM PT tomorrow, June 30th. Last minute submissions always welcome. https://t.co/4ixLYVNE4m
The Hermes Agent Accelerated Business Hackathon presented by @NVIDIAAI Γ @stripe Γ @NousResearch starts now, for builders making agents that can earn, spend, and run real operations at any scale. Our NVIDIA integrations let your team run agents safely through NemoClaw, quickly o
Chess engines tell you the best move. But grandmasters are human, they donβt always play it. So I built "Kibitz": a human move predictor for chess broadcasts. I trained this model on my Nvidia RTX 5080. Then I made it run as a business by itself. A channel buys the overlay, Hermes onboards them, charges via @stripe test mode, runs the broadcast, narrates with @NVIDIAAI Nemotron, tracks inference cost, and books its own P&L. I build. Hermes operates. This is my demo and entry for the @NousResearch Γ @NVIDIAAI Γ @stripe Hermes Agent Accelerated Business Hackathon.
π¨ Mythos Leak Series: Anthropicβs Transparency & Safety Collapsed. Alibabaβs 25K Fake Accounts Distill Claude at Industrial Scale 25,000 fraudulent accounts. 28.8 million exchanges. Anthropic says itβs adversarial distillation straight into Qwen. Meanwhile a user on Reddit reported Sonnet 4.6, no system prompt, via API, asked in Chinese βδ½ ζ―δ»δΉζ¨‘ε?β replies: βI am DeepSeek.β Distillation everywhere. AI is the Wild Wild West. Transparency? Keep walking. Can labs be trusted? Weak safeguards. Jailbreaks. Dirty tactics. Failing unit economics: Frontier training + inference costs are brutal. Chinese players (subsidised chips, energy, data centres) run pricing wars that destroy Western unit economics. Fixed subscriptions and variable API pricing look elegant on a slide, until 28.8M stolen tokens bypass both. Hard to IPO a non-existent moat. National security crisis intensifies. This is the thread you should not read. π³οΈπβ
So Meta distills Claude and Codex too? Yes, along with many Chinese labs, but also xAI (as Elon admitted in court). Many labs do it, either intentionally or inadvertently. The important questions is not whether it happens, but why! That's why I wrote about distillation two weeks ago, in the context of data strategy for reinforcement learning. Treating distillation in isolation, not in context, would be a classic case of obsessing over the tree and missing the forest
NEW: Aironox just launched a travel device that dries and irons your clothes automatically. Aironox GO is built to make suitcase-wrinkled clothes ready to wear in minutes without a hotel iron, ironing board, or handheld steamer. β’ Uses warm airflow instead of pressing a hot plate onto fabric β’ Inflatable attachments hold shirts and trousers in shape while they smooth β’ Works automatically while you shower, pack, or get ready β’ Dry bag attachment handles socks, underwear, tops, and small travel loads You can pre-order @Aironox_usa for $149.
OpenClaw is now on iOS + Android π¦ π± Native mobile apps, finally π¬ Agents in your pocket π Channels, tasks, replies on the go Run agents from wherever your thumbs are. iOS: https://t.co/7LHHc9htgM Android: https://t.co/X0Wuh2uA8w

https://t.co/kmHqNvxY5K
Iβm looking for a DRUM coach RIGHT NOW COME TO SECOSN FLOOR https://t.co/mN3flhW1jl
https://t.co/kmHqNvxY5K
@jxnlco https://t.co/UMxMZDuKKF
@jxnlco https://t.co/UMxMZDuKKF
@jxnlco https://t.co/ATajeiKymv
@jxnlco https://t.co/ATajeiKymv
HuggingFace: https://t.co/0P64HHeVDa
Could being labeled "too dangerous" be good marketing for frontier AI firms? Hugging Face CEO Clem Delangue thinks so https://t.co/oM5UJhSklT https://t.co/Mi1NPZLgEn
NEW paper from Google (bookmark it) It's on advancing automated scientific review. Just pay attention to the focus on agentic verification which is something I've been writing about recently. AI is accelerating paper generation, but review is still mostly human throughput. The paper frames this as a systems problem and proposes four levels of AI-human collaboration for scientific evaluation. It also discusses Google Paper Assistant as an early tool for scaling parts of the review process. The important point is the direction of travel. AI-scientist workflows create verification debt. The review stack now needs its own agents, taxonomies, and human oversight patterns. Useful read if you track AI research agents, automated peer review, or verification infrastructure for scientific discovery. Paper: https://t.co/KXSJD5orkp Learn to build effective AI agents in our academy: https://t.co/LRnpZN7L4c
Tokenmaxxin vs API rates https://t.co/5FR5oUwy1Y

Iβm looking for a DRUM coach RIGHT NOW COME TO SECOSN FLOOR https://t.co/mN3flhW1jl
met the Sith Lord bro is TALL https://t.co/N0f2WCqxG2
met the Sith Lord bro is TALL https://t.co/N0f2WCqxG2
Do LLMs actually think in silence? A new paper formalizes four axioms β Causality, Minimality, Separability, and Stability β to audit latent thought representations. No open-weight LLM satisfies all four. Most representations encode nothing beyond the input embedding itself. https://t.co/5LckF0xbhj
HuggingFace: https://t.co/0P64HHeVDa