Graph – turn your ChatGPT into AI-sorted RSS feeds
TLDR: Graph is a self-tuned AI filter that sits between you and the social web, so you can see the most relevant content in one place each day. --- Hey everyone. A long time ago I stumbled on this Aaron Swartz documentary, The Internet's Own Boy. It’s a conviction-builder, and it got me thinking a lot about RSS feeds and information agency. Now in a time of LLMs and embeddings, my friend David and I have been building something we call Graph, which flips the OG structure of RSS from following rigid sources to following a configurable list of topics, and letting this "interest graph" act as…
In plain words
Graph is an AI-powered content filter that replaces traditional RSS feeds by using topics instead of rigid sources. It pulls thousands of posts daily from across the social web, organizes them by topic using embeddings, and delivers the most relevant content in a personalized daily feed. Users configure their interests through ChatGPT or Claude, and the system learns what matters to them. Graph is designed for people seeking a more agency-driven approach to information consumption across social platforms.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
TLDR: Graph is a self-tuned AI filter that sits between you and the social web, so you can see the most relevant content in one place each day. --- Hey everyone. A long time ago I stumbled on this Aaron Swartz documentary, The Internet's Own Boy. It’s a conviction-builder, and it got me thinking a lot about RSS feeds and information agency. Now in a time of LLMs and embeddings, my friend David and I have been building something we call Graph, which flips the OG structure of RSS from following rigid sources to following a configurable list of topics, and letting this "interest graph" act as the main driver of what content will show up in a feed. Our MVP learns keyword interests from ChatGPT or Claude and uses them as embedded signal magnets for content pulled into our giant macro RSS. So, Graph retrieves thousands of posts from the social web each day, tags them with topics, then gives each user a unique rank-ordered feed of content that aligns with their work, hobby, or research interests. Please try it out and let us know what you think! Here are some quick notes and disclaimers. -- Sources -- As of now we have only about 1600 sources, which pull in about 3000 posts each day from places like Hacker News, Reddit, Product Hunt, YouTube, X, Substack, research journals, blogs, and traditional media. The content for now skews techy, but throw it some curve balls! If you tell ChatGPT you’re a farmer from Nebraska, it will give you tags that match that, and Graph’s content will show mostly posts about corn futures, trucks, livestock trading, and so on. It’s kind of neat to see how your LLM describes you and to immediately convert that into an autonomous scouting tool. -- Social -- You’ll also find some social feature side quests. You can follow friends and see what Graph recommends them. You can also see what topic overlap you have with other people, from the LLMs as well as optional connectors like Spotify, YouTube, Goodreads, and Letterboxd. We’re working on a few chat features as well. Soon we’ll have a Graph agent to riff with who can DM you leads to your current thing or make an optional intro to someone with overlapping goals or interests. -- Recommendations -- We’d really love your UX feedback and source recommendations. We’re not sure what Graph’s UI is going to evolve into, so we’re open to any and all ideas on how to make it the right balance of info-dense and engaging. Importantly, we could use some help on content source recs so Graph is more diverse in coverage. We hope to add a few hundred new sources and social accounts to pull into Graph each week, ideally, mainly from user suggestions. When logged in you can see an Info page with more background on why we’re building Graph and what features are on the way. This has been a fun project so far, and Graph is already showing David and me posts about [social web browsers] and [digital identity mapping] we would have never tracked down in our normal X or YouTube dives. You’ve probably felt seen when a friend sends you a super relevant link to something you’re working on that you wouldn’t have found otherwise. That’s what we’re going for. We’ll take any feedback and thoughts. Let us know how accurately Graph’s tags are snapping to your content or not, and thanks so much for checking things out! Link for signup: https://www.graph.cx/login
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, August 2025
the whole month →
- IS
I built the world's most impractical 1000-pixel display and anyone in the world can draw on it. It draws a single pixel at a time and takes 30-60 minutes to complete a single image. Anyone can participate in the project by voting for the next image to be drawn, and submitting images. https://kilopx.com/
Work · 2025 · benholmen.com
- KT
Kitten TTS is an open-source series of tiny and expressive text-to-speech models for on-device applications. We are excited to launch a preview of our smallest model, which is less than 25 MB. This model has 15M parameters. This release supports English text-to-speech applications in eight voices: four male and four female. The model is quantized to int8 + fp16, and it uses onnx for runtime. The model is designed to run literally anywhere eg. raspberry pi, low-end smartphones, wearables, browsers etc. No GPU required! We're releasing this to give early users a sense of the latency and voices…
Dev tools · 2025 · github.com
- IW
I was wondering how I can arrange objects along a spherical helix path, and read some articles on it. I ended up learning about parametric equations again, and make this visualization to document what I learned: https://visualrambling.space/moving-objects-in-3d/ feel free to visit and let me know what you think!
Life & fun · 2025 · visualrambling.space
- TC
For HTML Day 2025 [1], I made a web service that displays the current sky at your approximate location as a CSS gradient. Colours are simulated on-demand using atmospheric absorption and scattering coefficients. Updates every minute, without the use of client-side JavaScript. Source code and additional information is available on GitHub: https://github.com/dnlzro/horizon [1] https://html.energy/html-day/2025/index.html
Dev tools · 2025 · sky.dlazaro.ca