April 24, 2026
Dot clouds vs Doing stuff
Show HN: Atomic – Local-first, AI-augmented personal knowledge base
Open‑source note brain gets love, side‑eye, and dot‑cloud roasting
TLDR: Atomic launches as a private, open‑source AI wiki that auto‑organizes your notes and sources. HN loves the hustle but debates whether it beats “Obsidian + Claude,” mocks the pretty dot graphs, and asks for practical features like CSV clustering and more collaboration—because usefulness beats vibes.
Atomic promises a private, open‑source “second brain” that auto‑links your notes, saved articles, and web clips into a living wiki with citations. It’s self‑hosted (your data stays yours), searches by meaning, and even sends a daily AI briefing. The founder says he’s been “shipping like crazy,” rebuilding the iOS app with Android coming soon, while a Karpathy‑sparked wave of look‑alikes sets the stage for a showdown of personal AI wikis.
But Hacker News arrived with the spice. The top refrain: “How is this different from pointing Claude at Obsidian?” Translation: why not just aim an AI assistant at a popular note app and call it a day. Others roasted the ubiquitous galaxy map of dots—pretty, but “not useful beyond that,” said one skeptic. A reviewer noted it’s the second large‑language‑model (LLM) wiki on the front page today and wished for more teamwork in the space, calling it the “LLM curse” of everyone reinventing the same thing. Practical users asked for CSV uploads and simpler clustering, name‑dropping tools that pivoted to enterprise. Meanwhile, fans cheered the local‑first promise and inline citations, while critics demanded proof it actually helps them find answers faster. The verdict: promising brain, spicy brain‑trust review.
Key Points
- •Atomic is an open-source, local-first personal knowledge base available as a desktop app, headless server, and iOS app.
- •Content is organized as a knowledge graph where notes, articles, web clips, and feeds are automatically embedded, tagged, and linked.
- •Atomic generates wiki-style articles from tagged content with inline citations that update as new material is added.
- •Semantic search uses vector embeddings to find related ideas by meaning, not only by keywords.
- •New features include a daily AI-generated briefing and a mini-canvas that highlights source atoms within the knowledge graph.