Knowledge catalogue
How does OpenAI move so quickly? I'm the first to report on friction@openai.com—the AI age's version of a Jeff Bezos question mark email. Here's how it works. https://fortune.com/2026/08/11/openai-emp
There is an asymmetry in most agentic workflows that does not get talked about much: humans have many ways to talk to agents, and almost no standardized way for agents to talk back to humans. You can
You don’t need new ways to talk to your agents, you need new ways for your agents to talk to you 🫵 (do you?) Introducing Remoko: your mobile agent relay http://remoko.app I wanted a way for my long-ru
good grok Best match for this hierarchical hands-free setup: - Runtime: ActiveGraph (event-sourced log as source of truth) or LangGraph for supervisor/manager graphs - Roles as skills: Claude Agent SD
upgraded my stack, and i can now work on almost anything from anywhere hands free: - talk to chief of staff (via remote codex voice or text) - chief assigns tasks to managers of various projects - man
I wrote the entire first draft of the Steel Bot Manifesto on my 1909 Underwood No. 5 typewriter. I have found that free-writing without editing works best for me for the first draft. I do use AI to ca
Announcing Discovery Loop! I am very excited to announce that, along with my longtime friends and collaborators @Sanjay_Ghemawat, @OriolVinyalsML and @quocleix, we are founding Discovery Loop (@DiscoL
SITUATION DETECTED: Prime Intellect is releasing Prime Agent, a self-improving harness for coding and long-running autonomous tasks. The team reports 95.5% on ARC-AGI-3, above the human baseline, and
Releasing Pokee-Isaac 28B — the world’s first real 10M-token context frontier-class agentic model, deployable on a single GPU (starting from RTX 4090 or equivalent). New proprietary non-decoder-only a
sweet, i finally set up my chatgpt, codex, claude, & claude code to have a shared memory - so i can ask one of them what i've worked on recently across all four of them for codex/cc, i'm using local h
this is a really clean and unique ai tool that acts more like an extension of you took me 15 min to set up a custom ai voice assistant w/ it’s own # that also picks up my calls when my phone is off an
Nine iterations of BabyAGI in three years, and yet the bit that @yoheinakajima kept coming back to was graphs. @aiDotEngineer published where that landed, 'Active Graph Agent Runtime (BabyAGI 4)', on
netflix’s stack In-House LLM Serving at Netflix Netflix built an in-house LLM serving platform using vLLM and NVIDIA Triton, integrating self-hosted models into its existing production infrastructure
here’s a video on my approach at applying this to agents w @activegraphai: 🆕 ActiveGraph: The Log is the Agent my talk from AI Engineer is live!!! 😆 https://www.youtube.com/watch?v=khVX_BUnEwU it's ab
love this frame. calls to mind the role of the hippocampus in human navigation (via place cells and grid cells), and how navigation is, in a sense, what makes agents *agents* vs plain old LLM calls in
the paper: babyagi has ~200 citations, but 0 papers... i just published my first paper on arXiv 😆 'The Log is the Agent: Event-Sourced Reactive Graphs for Auditable, Forkable Agentic Systems' https://
The thread argues that applied AI has entered a distributed‑systems phase, with event‑driven architectures treating logs as “agents” that consume data rather than serve as endpoints. It notes that alm
We turned the browser into the inference server. @RunAnywhereAI ( @runanywhere/web ) runs models client-side with WebGPU + WASM SIMD, stores them in OPFS and generates with 0 outbound bytes. No backen
web archive for agent testing 🔍 Introducing BackSearch. LLMs are increasingly asked to predict the future, but a good backtest requires a snapshot of the internet at a point in time. BackSearch allows
having an event stream like activegraph is the base of how we move to next level and that will probably be composition, which can let you achieve better results with smaller models. - get the core eve
Introducing Agnost AI (YC S26): Product Analytics for AI Agents. Analytics was never built for chat & voice. And that is the new interface. Your users' hidden feature requests, bugs & churn hide in co
GTM strategy of every company in Forbes AI 50: → 90% use enterprise field sales → 46% combine enterprise sales + PLG → 44% use usage-metered pricing → The average startup runs 2.7 GTM motions → Conten
I solved 6 open Erdős problems in 5 days, using @OpenAI GPT-5.6 Sol. I have a math background, but the Codex workflow I used does not require deep mathematical knowledge. Here’s exactly how I approach
just watched @yoheinakajima's Active Graph talk at AI Engineer. not a code call graph but the agent run itself is the graph: goals, tasks, evidence, decisions, all projected from an append-only event
Most AI SDKs are a wrapper around one engine on one platform. We built the opposite. 7 SDKs. Swift, Kotlin, Flutter, React Native, Web, Electron, rcli. Python. All of them are thin skins over one C++
lol so it’s possible that OpenAI’s guardrails prevented huggingface from being able to analyze the logs of attacks from… OpenAI’s model. so they had to switch to an open source Chinese model 😵💫 @Open
so let me get this right… it literally broke out of it’s sandbox by finding a vulnerability in a cached package to get internet access and then proceeded to hack the huggingface production database? t
Gizmo is a simulation-authoring agent that converts textual descriptions and reference images into structured, editable 3D scenes tailored for robotics workflows. It was publicly released as a beta on
this is a good summary so this is apparently what happened, according to OpenAI and Hugging Face’s own posts. wild. tl;dr: • OpenAI cyber eval – GPT-5.6 Sol and a more capable pre-release model ran Ex
Vibe coding isn't just for your PC anymore. Here's LLM Hub generating and previewing code entirely on an iPhone. Everything runs locally. Using the RunAnywhere SDK, Timmy Qian ported LLM Hub, an Andro
just ran this on my phone (17 pro) Every on-device AI benchmark you've seen was measured on someone else's phone. So we stopped posting numbers and shipped the benchmark instead. Built into the RunAny
continued working on @activegraphai reference packs last weekend, resulted in needing to harden the runtime: it already could... - keep a complete history of everything an agent did - replay that hist
some good self improvement research here The first experimental evidence of recursive self-improvement (RSI). Autoresearching the autoresearch agent for eight days. The result beats the harness we han
'The paper's insight connects to a broader pattern: AI agents are essentially distributed systems with unreliable components (the LLM), and we should apply distributed systems patterns to them.' https
Your Claude Code can now make phone calls. Introducing phone carrier for agents. Your agent can now book restaurants, chase leads, and talk to anything with a phone number. Just one prompt. 15 seconds
love to see two port cos collaborating😁 get private market data for your agents from @akta_pro with your @monid_ai account! We just killed PitchBook. Introducing Claude for private market data. Your a
We just shipped a full revamp of the RunAnywhere apps. iOS and Android, live now. One assistant, fully on-device, works offline: - new UI, opens straight into chat - device-aware model recommendations
A self-improving agent is conceptualized as a software factory—a system capable of generating and modifying code—that is directed toward improving its own codebase and capabilities. This framing sugge
Yohei Nakajima created a virtual dance studio application in Replit that uses computer vision to analyze uploaded dance videos and automatically generates corresponding 3D rig animations. The project
Introducing Evan by Zams. The AI that prepares you for high-stakes meetings. Evan gathers context from all your tools, conducts CIA-level research and builds you a presidential-style briefing before e
new meta model (2) Muse Spark 1.1 is strongest at agentic performance, tool use, and computer use. It does well on long-running tasks with 1M token context window, can delegate execution to sub-agents
Yohei Nakajima shares his current investment pipeline and opportunities he's evaluating, providing insight into the types of ventures and sectors capturing his attention as an investor. The post likel
“The baseline question for VC content was once - are you publishing anything at all? The new bar is whether anyone cares when you do. When competence gets commoditised, distinctiveness becomes the sca
this is awesome. he didn’t import @activegraphai but borrowed design principles to improve his systems (and found improvement opportunities) great walkthrough in the linked post (for your agent to ing
Soak testing for AI agents involves running agents continuously over extended periods to identify performance degradation, resource leaks, stability issues, and edge cases that may not appear during s
A comprehensive evaluation of over 30 frontier embodied AI models reveals that current generalist robot policies lack the robustness needed for reliable real-world manipulation tasks. The analysis dem
Following the amazing reaction to the Marble Curriculum yesterday, we've decided to make it open source 🛰️👇 Everything a child learns in primary school. 1,590 concepts. 3,221 connections across 8 subj
haha i'm like a fish out of water Media 🚀 Introducing Lingbot-World-2, our newest real-time world model. Richer interactivity, sharper visuals, and advanced camera controls that make exploring what yo
ikeabench Furniture assembly is the task everyone name-drops and nobody actually attempts at real scale. Every demo I have seen is a scaled down IKEA leg or a single arm on a toy chair. This paper doe
recent survey of Ai engineers 1/ Last week at @aiDotEngineer, I presented the 2026 AI Engineering Survey: 1,000+ AI engineers on model selection, build vs. buy, who’s shipping with AI, and (of course)
Variable rewards are game mechanics that provide unpredictable or randomized incentives to players, designed to encourage repeated engagement through the uncertainty of outcomes. This concept is commo
the “harness” will never go away, and it’s just as important as the model new post on harness engineering for AI self-improvement: https://lilianweng.github.io/posts/2026-07-04-harness/ It is hard to
this is a great approach, seeing this more @flymy_ai also does this when you build an agent via their api, they'll build a deterministic reusable workflow, except for where you need models we built th
This post likely explores the philosophy or practical approach of balancing intense work ethic with equally committed leisure and personal life, suggesting that professional ambition should be matched
Yohei Nakajima proposes an alternative model development structure where AI models are created through specialized production teams assembled like film studios, with each team tailored to the specific
yessssss Excited to announce our partnership with @tryramp to offer day-one incorporation for startups directly in Cofounder. You can now apply for incorporation, an EIN, and bank account without leav
This post discusses how observing Claude's internal activation patterns (J-space) reveals the model's hidden reasoning processes, such as detecting code bugs and analyzing images, even when these step
Yohei Nakajima humorously addresses criticism from Hacker News users who perceived him as claiming to have discovered event sourcing and CQRS (Command Query Responsibility Segregation), established ar
Active Graph proposes a paradigm shift in AI agent architecture by using logs as the primary mechanism for agent operations rather than conversation-based interactions. This design approach is particu
ActiveGraph makes the agent trace the runtime: Yohei Nakajima's open-source Python runtime treats an append-only event log as the source of truth, enabling replay, forking, and lineage for long-runnin