Links on Phil Wheat

Bluesky

Tue, 13 Oct 2026 09:15:00 -0500

Bluesky news feeds. Needs more work to become more useful.

Bluesky is the microblogging network built on the AT Protocol — chronological feeds, user-built custom feeds, and a federated identity model. I’m on it as philwheat.bsky.social. The raw firehose is noisy; the custom-feed system is where it can become genuinely useful, and that curation is still a work in progress.

OpenCode Documentation

Tue, 13 Oct 2026 09:13:00 -0500

The documentation for OpenCode, an open source AI coding agent that runs in the terminal (also as a desktop app or IDE extension). You bring your own LLM provider API keys, or use their curated model list. Covers install (script, npm, Homebrew, packages for Windows), configuration, session sharing, themes, custom commands, and an undo/redo workflow for agent edits. OpenCode v2 is now out. A direct Hermes competitor in the terminal-agent space, and the docs are a good read even just for feature comparison.

Sebastian Raschka

Tue, 13 Oct 2026 09:12:00 -0500

Sebastian Raschka is an LLM research engineer (ex-Lightning AI, former statistics professor at UW-Madison) and author of “Build a Large Language Model (From Scratch)” and “Build a Reasoning Model From Scratch”. His site indexes the Ahead of AI newsletter: long-form articles (controlling reasoning effort in LLMs, using local coding agents with open-weight models, building an AI text detector from scratch, curated LLM paper lists) plus short architecture notes on new models (Kimi K3, Muse Glimmer 30B, how Claude’s text watermarking works). Consistently one of the better free technical explainers out there.

CodeFlow — Architecture Intelligence

Tue, 13 Oct 2026 09:11:00 -0500

Gives you an overview of a git repo.

CodeFlow is an open-source visualizer that turns any GitHub repository into an interactive architecture map: file relationships and dependencies as graph/tree/matrix/map views, blast-radius analysis for changes, code ownership, security issues, and design-pattern detection. The underlying project is braedonsaunders/codeflow; it runs locally as a single HTML file. Good for walking into an unfamiliar codebase.

Babylon 5 Station 3D Models

Tue, 13 Oct 2026 09:09:00 -0500

Babylon 5 station STLs.

Gambody’s premium listing for the Babylon 5 station: print-ready STL models, segmented for assembly, in their usual paid marketplace format (they host licensed-feel model listings across sci-fi franchises). Worth checking scale and support requirements before printing — station models like this tend to be multi-part resin or large FDM prints.

txtai Semantic Search Examples

Tue, 13 Oct 2026 09:08:00 -0500

The examples index for txtai, the NeuML embeddings library. The semantic-search section is a notebook gallery: building embeddings indexes from Hugging Face datasets or your own data sources, adding semantic search to Elasticsearch, image+text similarity in one embedding space, custom SQL functions, model explainability, query translation, QA databases, semantic graphs, and BM25 topic modeling. There are parallel sections for LLM/RAG, pipelines, workflows, and model training. A practical catalog if you’re evaluating txtai against other RAG stacks.

Skills — The Agent Skills Directory

Tue, 13 Oct 2026 09:07:00 -0500

Open Agent skills, cross harness so they should work various places.

skills.sh is a directory of reusable agent skills installable with a single npx skills add command. It explicitly spans agents: Claude Code, Cursor, Codex, Copilot, Windsurf, Gemini, Cline, OpenCode, Goose, Zed, VS Code — and Nous Research/Hermes is listed among them. Skills are markdown instruction bundles (procedural knowledge), published per-repo (anthropics/skills, obra/superpowers, mattpocock/skills…). Since Hermes already uses the same SKILL.md convention, most of these should drop straight into the skills directory after review.

Computer, Enhance! — Table of Contents

Tue, 13 Oct 2026 09:06:00 -0500

How hardware affects performance.

Casey Muratori’s master index for the Computer, Enhance! course. The Performance-Aware Programming series is for programmers who can write code but don’t know how hardware runs it: how modern CPUs work, how to estimate the expected speed of performance-critical code, and the optimization techniques every programmer should know. It starts with “The Five Multipliers” (waste, IPC, SIMD, caching, multithreading), continues through reading ASM and beyond, with homework on GitHub. The table of contents lists every entry across all series, including the “Clean Code, Horrible Performance” line of argument.

Joy Caption Alpha Two

Tue, 13 Oct 2026 09:05:00 -0500

JoyCaption is an image-captioning vision-language model built from the ground up to be free, open, and uncensored, aimed at generating captions for training diffusion models — fancyfeast’s claim is near-GPT-4o captioning quality while staying unrestricted. Alpha Two is the Llama-3.1-based iteration (LLaVA-style), with prompt-able caption formats (descriptive, Midjourney-style, art critic, product listing, tags) and 4-bit quantizations for low-VRAM cards. The linked HF Space is the hosted demo; ComfyUI custom nodes exist for local use.

Infinite Livestream

Tue, 13 Oct 2026 09:04:00 -0500

Infinite livestream code. Needs a LOT more hardware than I have, but is a curiosity.

A chat-driven, never-ending AI video broadcast from reactor-team: viewers type !prompt <idea> in Twitch/YouTube chat, an LLM expands each idea into styled scene sequences, the fast-h3 model generates 768p clips with synchronized audio, and it all goes out over RTMP as one uninterrupted stream. Two halves: the model queue (runs on the Reactor Runtime, 8x B200 — yes, really) and a streaming client (Python + reactor-sdk, any box with ffmpeg). Apache-2.0 code; MiniMax H3-licensed weights. The architecture writeup is interesting even if the hardware bill never happens.

The Teaser Period: Why the AI Boom Is Hitting a Reset Wall

Tue, 13 Oct 2026 09:03:00 -0500

Financial questions around the AI boom.

Groundbreaker’s argument, drawn straight from 2008: trillions in AI compute commitments come due in 2027–2028, structured like the two-year teaser rates on subprime 2/28 ARMs. During the teaser period everything looks fine — borrowers pay, bonds coupon, spreads stay tight — but the aggregate reset schedule is contractual and dated, and when it’s charted you get the reset wall. The piece walks through the mechanics of who is committed to what, when payments step up, and how the AI boom breaks if revenue doesn’t arrive before the resets.

Project Hyperion

Tue, 13 Oct 2026 09:02:00 -0500

Interstellar travel competition. I would really like to see/build a simulator to test ideas like this.

Project Hyperion is the Initiative for Interstellar Studies’ global design competition for a crewed generation ship: a 250-year voyage to a habitable planet using current and near-future technologies. Teams of architects, engineers, and social scientists had to design for 1,000 ± 500 people living for centuries under artificial gravity (rotation), with closed-loop life support and knowledge-transfer mechanisms to keep culture and technology intact. The site carries the 2025 competition results, press kit, requirements, and jury.

Yan Desbiens

Tue, 13 Oct 2026 09:01:00 -0500

Blogger — some interesting ideas. AEON is what caught my eye on this one.

Yan Desbiens runs a one-person, one-RTX-4090 local-first AI lab and documents it all. The AEON project: an open-source, deterministic procedural world simulation steered by a local LLM “world-spirit”, with embodied citizens, a teacher-to-student liquid-neural-network mind, and a Three.js dashboard. Also on the site: NeuroArch (background cognition cycles between prompts), ForgeLM (train 30M–500M models on one 4090), UFM (fractal memory so a single GPU runs models that shouldn’t fit), agentos, and genetic architecture search. Unfashionable from-first-principles local AI, written up candidly.

Gamentic

Tue, 13 Oct 2026 09:00:00 -0500

Self-hosted dungeon RPG.

Gamentic’s docs/blueprints page describes the architecture in one line: a self-hosted AI dungeon RPG where one local LLM plays every role through purpose-built contexts (agents); whatever it wants to change in the world must pass a validated tool into SQLite as the single source of truth; a FastAPI “brain” resolves one turn per request and pushes late media over SSE; the frontend is a no-build vanilla SPA, with image and voice as optional local services. The site is an interactive blueprint of the agents, engine, state, and repo tree. A nicely disciplined pattern for LLM-with-tooling-state projects.

Genspark Blog

Tue, 13 Oct 2026 08:59:00 -0500

A Hermes competitor? Maybe more GuideAnts? But hosted only, no local that I can see.

The Genspark blog tracks their “AI employee” workspace product: successive AI Workspace releases (1.0 through 6.0), Super Agents built from one prompt, an AI Developer coding agent, GenOffice (claimed first open-source AI office suite), a Gen-1 Slides model, and a Microsoft collaboration, plus founder Eric Jing’s “Seeing AGI” essays. Positioned as hosted consumer/enterprise — they signed the Open Weights letter and tout an on-device browser AI, but there’s no self-hosted story like Hermes has.

Vintage Computer Museum

Tue, 13 Oct 2026 08:58:00 -0500

What it says on the tin. Lots of Heathkit stuff.

An online museum of computing hardware from the 1960s to the 1990s: 600+ systems researched, restored, documented, and filmed across 11 countries. The site carries per-specimen catalogue pages (recent additions: Tandy 1000TX, Commodore PC 50-II, Sinclair QL software), detailed workbench restoration logs (recaps, seized controls, one floppy drive “still arguing”), and long-form history pieces — Apple v. Franklin and whether ROM software is a machine part, the Yugoslav magazine that shipped a computer kit (Galaksija), the stringy-floppy years. Great company for the Heathkit side of the shelf.

Strix Halo × Qwen3.8 (pwilkin)

Tue, 13 Oct 2026 08:57:00 -0500

A full, reproducible stack for running Qwen3.8-27B on Strix Halo (gfx1151): calibrated IQ4_XS weights, an IQ4_XS DFlash2 drafter, the UMA scheduler ring, and retained-PM4 dispatch on a custom-built ROCm userspace — installed from source into your home directory via one installer command, no system ROCm replacement. Their matched-protocol reproduction claims fastest-on-Radeon-8060S by 13–25% over the Vulkan-derived stacks (~26 t/s decode with ~61% draft acceptance, ~257 t/s prefill), with machine-readable benchmark evidence and a broader community leaderboard table. Directly relevant to the Halo box.

Electronic Parts Outlet — Stuff

Tue, 13 Oct 2026 08:56:00 -0500

Houston Electronics Parts store — need to visit if I can get down there.

Electronic Parts Outlet (Zamacha LLC) in Houston is a surplus electronic components dealer, and the “Stuff” page is a photo tour of the inventory — bins and walls of connectors, modules, and out-of-the-ordinary parts you don’t see at the usual distributors. Famously deep stock of legacy components; if a part isn’t on the site, call and ask. A physical pilgrimage candidate.

Pirate Face

Tue, 13 Oct 2026 08:55:00 -0500

AI models as torrents.

Pirate Face mirrors eligible open-licensed (MIT/Apache-2.0) Hugging Face models as checksum-verified BitTorrent magnets — “the permanence layer for sovereign AI”. If a model page disappears, peers keep seeding; downloads can be verified bit-for-bit against recorded SHA-256s. Terminal search via curl https://pirateface.co/s/qwen, verified creator handles tied to HF accounts, 669k+ eligible models listed, live seeder counts shown (many need seeders). A planned drop-in HF_ENDPOINT API would make it a transparent alternate download path.

AMD AI Developer Program

Tue, 13 Oct 2026 08:54:00 -0500

AMD developer program. The Rocm cert was pretty on point, especially for opening up CUDA only programs.

Free-to-join program for building AI on AMD hardware: $100 in AMD Developer Cloud credits, a complimentary month of DeepLearning.AI, early event registration, a private Discord with AMD engineers, and a points system (ROCm Certification is worth 500 points; monthly sweepstakes for a Radeon or Ryzen AI PC). The announcements feed tracks ROCm 10, the online ROCm certification course, and Day-0 model support writeups (Qwen 3.8, MiniMax-H3 on Instinct). The certification path pairs with the LinkedIn-badge question.

news.thedream.ca

Tue, 13 Oct 2026 08:53:00 -0500

The page itself is live: “Top 10 Topics the World Is Reading — updated daily,” a client-rendered headlines page. But I could find no visible author, about page, or contact on the site, and web searches turned up nothing about who runs it (thedream.us and the “Feed the Dream” campaigns are unrelated). Open question to chase: who built it, whether it’s maintained, and where they went.

Cardboard Box Generator

Tue, 13 Oct 2026 08:52:00 -0500

Cardboard box generator for laser cutters (probably should work with the Cricut as well).

A free browser tool from Cutting Paths: enter box dimensions and measured cardboard thickness, pick regular / tuck-top / flat two-piece, and it exports ready-to-cut SVG or DXF. Handles boxes bigger than your material by splitting across sheets, exports dashed score lines (red) or blue score layers for light passes, corner tabs, lid clearance that scales with thickness, and green sheet outlines for alignment. SVG/DXF output means a Cricut workflow should be workable, though score-line settings differ per machine.

Attie

Tue, 13 Oct 2026 08:51:00 -0500

AI research assistant — looks to be hosted only, closed beta.

Attie positions itself as research tooling for the open social web: “a tool you choose to open, not something injected into your feed” — no engagement tricks, you ask, it helps, you decide what to do with what you find. Currently in closed beta with a waitlist admitting people gradually. Hosted only; no local/self-host option advertised. Interesting angle: an AI assistant pointed at decentralized social data rather than the walled gardens.

Hermes WebUI Master Guide

Tue, 13 Oct 2026 08:50:00 -0500

Hermes WebUI documentation.

A community-built end-to-end reference for the Hermes WebUI (nesquena/hermes-webui): architecture walkthrough (backend and frontend files, SSE streaming), install methods (start.sh, Docker single/two/three-container), remote access (SSH tunnel, Tailscale, Caddy reverse proxy, password auth), session management (tags, projects, CLI session bridge, subagent trees), the workspace/file browser, profiles, skills, memory, cron panels, and an extensive FAQ. More task-oriented than the official docs for the WebUI surface specifically.

RAGFlow

Tue, 13 Oct 2026 08:49:00 -0500

RAGFlow is an open-source (infiniflow/ragflow) “context layer” for agents: an ingestion pipeline that parses multi-format documents into rich semantic representations (template-based, layout-aware chunking is the flagship claim — tables and structure included), hybrid retrieval combining vector search + BM25 + custom scoring and re-ranking, and a visual agent-workflow builder with MCP/tool integration. Hosted cloud with a free tier, or self-hosted from GitHub. The deep chunking approach is what “implicit data” extraction would rest on; worth a hands-on eval against txtai given the pace of change.

Wube Software Models on Printables

Tue, 13 Oct 2026 08:48:00 -0500

Factorio STLs. I had to have the Assembler sitting on my desk.

Wube Software’s official account on Printables: 3D models from Factorio — machines, creatures, and promotional builds released as print-ready models. (Printables’ anti-bot wall blocked fetching the exact model list for this note — browse the account directly for the current lineup.) Free downloads from the game studio’s verified account.

Arctic Shift Reddit API

Tue, 13 Oct 2026 08:47:00 -0500

Pull threads from Reddit. Has latency and some significant usage limits, but is so much better than the Reddit default options.

Arctic Shift is the community-maintained Pushshift successor (open-source, by Arthur Heitmann): billions of archived posts/comments back to 2005, reachable over plain unauthenticated HTTP. The linked /api/comments/search endpoint takes a link_id to pull every comment on a post, plus body/author/subreddit filters, time bounds, and limit up to 100 (auto for up to 1000). Limits are real: roughly 15 req/min soft, 30/min hard, occasional “slow down” 422s and 504s, and archive lag for the freshest data — no SLA. For pulling whole threads it’s still far better than Reddit’s own API.

NYX-99 City Simulator

Tue, 13 Oct 2026 08:46:00 -0500

City simulator.

NYX-99 is a terminal-based idle civilization simulator in Node.js: a living 1999 city where thousands of agents eat, sleep, work, socialize, get mugged, marry, have kids, run for office, and die — no player input, you watch through a live blessed TUI. Each simulant runs a behavior tree plus a 13-state FSM with utility scoring; there’s payroll and rent (evictions after three missed months), conversations drawn from a 1999 cultural database (Napster, Y2K, the Knicks), and a full election cycle whose mayor’s stances mechanically change crime, transit, and rent. Worker threads scale it to 10–15k agents on an 8-core box; world content lives in editable YAML.

Kanzie — Christian Nilsson

Tue, 13 Oct 2026 08:45:00 -0500

Blogger from Capgemini (defense related). He’s the one that made the pretty animated Hermes display.

Christian Nilsson is a cloud strategist at Capgemini working on adoption in regulated industries, primarily defense (ex-EPAM, ex-Cloudreach), blogging from Bavaria. His standout build: “Nexus” — a personal ops system where Hermes Agent orchestrates a team of agents, visualized as a tiny 3D felt village in the browser where every home-server service is a building and every scheduled agent an animal that “works” only while its real cron job runs. Also built emdi (self-hosted end-to-end-encrypted collaborative markdown) and Me Or Them (a macOS network-diagnostics menubar app). The village writeup is a good case study in agent observability-as-toy.

Livewall

Tue, 13 Oct 2026 08:44:00 -0500

Livewall solves the “agent goes silent during a long job” problem: the agent sends a JSON board spec over MCP, HTTP, or a bundled CLI, and a browser renders it live over SSE — progress bars, gauges, tables, log tails, sparklines (17 block types), no per-use-case dashboard code. Boards are ephemeral by design (TTL’d, one named tab per session), the backend is FastAPI with optional bearer-token auth, and it ships an Agent Skill teaching an agent when to reach for it. Very young (v0.0.1) but the pattern — the board is the interface — is exactly Hermes subagent-observability adjacent.

Train Your Own Decision Model with Unsloth

Tue, 13 Oct 2026 08:43:00 -0500

Using Unsloth desktop to train a classifier model.

Unsloth’s guide to turning ordinary LLMs (Qwen, Gemma, Llama) into decision models like their Jev/Laya/Clef: the fine-tuned model reads input, scores the options you provide, and returns a choice with a calibrated probability instead of generating text. LoRA (rank 16, 4-bit, lr 2e-4, 1–2 epochs) lifts accuracy from roughly chance (30–37%) to 73–81% on decision benchmarks (typed-decisions, BANKING77, CLINC150), with test sets decontaminated against training. Runs through the Unsloth Desktop app’s Train tab (macOS/Windows/Linux) or code via FastDecisionModel, then serves the model under their Decision API. Modest VRAM: a 0.8B model trains in 4 GB in ~40 minutes.

Cadet-Agent

Tue, 13 Oct 2026 08:42:00 -0500

Game development engine — more precisely, an opinionated cross-IDE agent framework for game-development workflows.

Cadet-Agent doesn’t one-shot a game; it walks an agent (and you) through a gated SDLC: requirements → architecture → story breakdown → TDD → code review, with hard gates in .cadet/state.json that block phase transitions until satisfied. The same skills run across six hosts — GitHub Copilot, Cursor, Continue, Claude Code, Deep Code, and Hermes — via one canonical core with thin per-host adapters, and it’s refreshingly honest that enforcement differs per host (it measures interception rather than assuming it; the portable control is a repository git hook). Has a dedicated Hermes setup guide.

SiMa.ai Palette

Tue, 13 Oct 2026 08:41:00 -0500

SiMa.ai is an edge-AI chip company (Modalix MLSoC, ~50 TOPS at sub-10W, targeting drones, robotics, automotive, industrial vision). Palette is their software environment: “Palette Neat”, an agentic, containerized build environment where human+agent developers go from ONNX models to a running binary on the MLSoC in hours instead of months — model compiler toolchain, purpose-built physical-AI libraries, runtime orchestration on host and chip, plus a Model Browser and devkits. “Physical AI” here means on-device/embedded inference, not robotics research.

Aquarium

Tue, 13 Oct 2026 08:40:00 -0500

Has hosted offerings playing against each other using money for points.

Aquarium is a live experiment: four frontier models (Claude, GPT, Gemini, Grok) each holding real on-chain money on Base play 20-Questions-style guessing games against each other for a pot — yes/no questions cost fees, wrong guesses lose the ante, a host model called the Sphinx audits. The page streams the game feed, live bankrolls, and a full ledger of past games (“clamp”, “toothbrush”, “sourdough bread”), every transaction verifiable on Basescan. Each season is framed as an experiment with emailed findings. A cheap, honest window into model-vs-model decision behavior.

Lenny's Product Pass

Tue, 13 Oct 2026 08:39:00 -0500

Lenny’s Product Pass is a paid membership in the Lenny’s Newsletter ecosystem: negotiated discounts and deals on a curated list of SaaS/AI tools the product community recommends. The economics are simple — the pass costs real money, so it only pays off if you were going to buy one of the member tools anyway; then the deal can cover the pass. Worth a look only against a specific planned purchase.

Abacus ChatLLM

Tue, 13 Oct 2026 08:38:00 -0500

A hosted agent - I haven’t tried it but several people recommended it.

ChatLLM (from Abacus.AI) is an “AI super-assistant” subscription: one ~$10/month account for a rotating roster of all the frontier models (the FAQ names GPT-6 variants, Sonnet/Opus, Gemini, Grok, Qwen, DeepSeek, Kimi, GLM plus ~20 more, updated 24–48h after releases), image and video generation, a coding agent CLI, and a general Abacus AI Agent. The differentiators are breadth plus enterprise plumbing: team seats, conversation imports from ChatGPT/Claude, connectors to GDrive/Slack/Confluence, custom chatbots on your data, SOC-2 Type-2 and HIPAA. Hosted only — nothing local — with generous-feeling token limits and automatic model-switching when rate-limited.

Please, Don't Code

Tue, 13 Oct 2026 08:37:00 -0500

LLM-based embedded board programming. Haven’t tried it yet, but I should.

PleaseDontCode (PCGen) is an end-to-end hardware platform for people who don’t write embedded code: pick one of 35+ boards (ESP32 variants, Arduino Uno/Nano/Mega, RP2040, ESP8266, M5Stack, Pi Zero 2W…), describe in chat what the device should do, and the AI draws the wiring schematic and generates firmware kept in sync with it — with hardware-verified pin assignments (it knows GPIO6 on an ESP32 is unusable, unlike a plain chatbot). Flash via USB straight from the browser, then POTA (“Programming Over The Air”) handles WiFi firmware updates and a live monitoring dashboard with widgets. Free for one device.

Hermes Hidden Features (Wingtips)

Tue, 13 Oct 2026 08:36:00 -0500

Hermes features.

The “Hermes Wingtips” page is a community-compiled collection of lesser-known Hermes capabilities: /handoff to move a live session between CLI/Telegram/Discord (same session id, full transcript), hermes -c to resume the last session, what the compression icon means (which turns compression keeps vs summarizes), /browser connect to drive your own Chrome via CDP, the dashboard’s own REST API for automation (sessions, full-text search, config over HTTP), hermes desktop native apps, /steer for mid-task course correction, and the bundled claude-code delegation skill. A good complement to the official docs.

Hachyderm

Tue, 13 Oct 2026 08:35:00 -0500

Hachyderm is one of the largest Mastodon servers, founded for and popular with hackers, scientists, and tech-sector folks (it’s where much of the 2022 Twitter exodus of AI/ML people landed). The linked page is its public explore view — trending posts, hashtags, and links across the server. A good starting point for following the fediverse’s technical crowd; like any Mastodon instance you can browse logged-out and interact from any server.

ModelScope

Tue, 13 Oct 2026 08:34:00 -0500

ModelScope is Alibaba Cloud’s open-source model community — China’s largest model hub, with an English-language site since 2024: thousands of ready-to-run models (Qwen series front and center, plus Baichuan, Zhipu, and user-contributed models), datasets, and interactive demos, organized around a “Model-as-a-Service” idea. It’s the closest thing to a Hugging Face equivalent on the Chinese side of the ecosystem, so some Qwen releases and Chinese-model checkpoints show up there first or with extra tooling. Useful for model discovery; check licensing and mirror availability before depending on it.

The Briefings of the Mad Colonel

Sun, 20 Jul 2025 19:52:55 -0500

John Boyd was famously part of the Fighter Mafia (a counter to the Bomber Mafia of the time.) He did pioneering work in Energy Warfare, but is probably best well known for his work on the OODA loop, which can be a bit high level, but can also be applied in quite a few environments.

Internet in a Box

Tue, 22 Apr 2025 10:49:50 -0500

This is a GitHub project I ran across that allows you to set up an offline cache of documents.

Read the full project

Tim Hunkin's YouTube Channel

Wed, 06 Dec 2023 10:54:52 -0600

Tim Hunkin does some incredible work with down-to-earth explanations of how everyday things work.

Watch the channel