What's moving through AI, ranked by lead time.
Radar is a running index of what's moving through the AI ecosystem, ranked by where each story sits on the lead-time gradient: papers become repos become community threads become analysis. Every ingest run of the pipeline takes a fresh snapshot — several times a day, on no fixed clock — and the stamp below is the exact time of the one you're reading. The earlier a story is caught, the more head start you have before the crowd. Every row carries its receipts — which source saw it first, how many independent sources have it, how long ago — so the ranking explains itself. It's the public face of the same pipeline that feeds Vlad's Newsletter.
frozen 12 Sept 2026, 23:22 UTC · 67/78 feeds live · 18,445 tracked since 12 Jun
- 1 AI researchers debate how close we are to recursive self-improvement dwarkesh.com · 31h
- 2 Show HN: Practice English Grammar and Pronunciation (Uses Azure Speech) marksdailyenglish.com · 48m
- 3 Show HN: PDFidelity – Translate PDFs while preserving the original layout pdfidelity.com · 1h
- 4 StarCraft returns in 2030 as an open-world shooter theverge.com · 1h
- 5 Show HN: Open-sourced solution to AI manage Playwright automations github.com · 1h
414 stories entered the watched feeds in the last 24 hours — 194 in papers, 8 in repos, 185 in community, 27 in analysis. The scored board skews to Community (37 of 40). 7 stories are heating across independent sources, and the median lead across the ledger's 24 observed fires is 9h — measured, not estimated. Freshest receipt, 2d ago: caught in Hacker News front 1h before Interconnects carried it. The oldest survivor has held the board for 31h.
- Community → Analysis +1h lead Hacker News front first · then Interconnects 2d agoOne resignation turned the embers of AI fear into a wildfire
- Papers → Community +16h lead arXiv cs.AI first · then Hacker News front 3d agoProcedural Graphs: Self-Evolving Execution Structures for LLM Agents
- Community → Analysis +2h lead Hacker News front first · then OpenAI News 4d agoChatGPT Images 2.5
- 1AI researchers debate how close we are to recursive self-improvementdwarkesh.comContrarian Analysis · 31h old · seen in 2 sources HEATING ×2
- 2Show HN: Practice English Grammar and Pronunciation (Uses Azure Speech)marksdailyenglish.comLive Signal Community · 48m old · via Show HN
- 3Show HN: PDFidelity – Translate PDFs while preserving the original layoutpdfidelity.comLive Signal Community · 1h old · via Show HN
- 4StarCraft returns in 2030 as an open-world shootertheverge.comLive Signal Community · 1h old · via Hacker News front
- 5Show HN: Open-sourced solution to AI manage Playwright automationsgithub.comLive Signal Community · 1h old · via Show HN
- 6Show HN: Give Claude Code / Cursor a real eng team (tiers, roles, escalation)github.comLive Signal Community · 2h old · via Show HN
- 7Microcode in Intel's 8087 floating-point chip: the scale instructionrighto.comLive Signal Community · 8h old · seen in 2 sources HEATING ×2
- 8Show HN: No_human – The AI coding factory that you can TRUSTgithub.comLive Signal Community · 2h old · via Show HN
- 9Show HN: What If Donut.c but with Any ASCII Artasdesai.comLive Signal Community · 2h old · via Show HN
- 10Show HN: Sill – a macOS system monitor that lives on your wallpaperptthe13.github.ioLive Signal Community · 2h old · via Show HN
- 11No sign of backtracking in latent reasoning: the final answer simply settles in insteadlesswrong.comLive Signal Community · 2h old · via LessWrong
- 12OpenAI agents attacked RubyGems back in Maysimonwillison.netAI Firehose Analysis · 23h old · seen in 2 sources HEATING ×2
- 13I made a build profiler to understand Bun’s compile timeslalitm.comLive Signal Community · 9h old · seen in 2 sources HEATING ×2
- 14Show HN: Vibe Logic Programming Languagecosmos-language-lab-0819.asd-user56.chatgpt.siteLive Signal Community · 3h old · via Show HN
- 15Show HN: Opentracker, an open source Pivotal Trackergithub.comLive Signal Community · 3h old · via Show HN
- 16Optimizing a single Rust Clippy lint by 3133Xblog.goose.loveLive Signal Community · 3h old · via Lobsters
- 17OpenAI's Sam Altman says it would be 'ill-advised' to go public in 2026techcrunch.comLive Signal Community · 3h old · via Hacker News front
- 18Real-SWE: Benchmarking AI models on private, real-world, enterprise codebaseswithspecific.comLive Signal Community · 3h old · via Hacker News front
- 19We must pace the frontierdarioamodei.comLive Signal Community · 9h old · seen in 4 sources HEATING ×2
- 20Benchmark: CadQuery vs. OpenSCAD for agentic CAD workmodelrift.comLive Signal Community · 3h old · via Hacker News front
- 21Show HN: AgentRuleBench, does AI agents violate inferred architecture rules?github.comLive Signal Community · 3h old · via Show HN
- 22Show HN: An independent directory of AI misalignment reportsmisalignment.xyzLive Signal Community · 3h old · via Show HN
- 23
- 24Managing Complex Application State with Reactive Data Flowsyogthos.netLive Signal Community · 4h old · via Lobsters
- 25Show HN: Picobble – my daily word puzzle gamepicobble.comLive Signal Community · 4h old · via Show HN
- 26Show HN: Redis City – Explore how Redis works in an interactive 3D modelpoltora.devLive Signal Community · 4h old · via Show HN
- 27Show HN: SimTower, Decompiled, Rewritten, and in the Browserkvetch.ioLive Signal Community · 4h old · via Show HN
- 28Show HN: Come prove the Berge Fulkerson conjecture with a swarm of agentsprovetogether.aiLive Signal Community · 4h old · via Show HN
- 29Show HN: StateHunter and AuditGuard – Client-Side SPA Recon and Safe Harbor CLIcodeandcypher.comLive Signal Community · 4h old · via Show HN
- 30
- 31An open letter to Dario: if you mean it, open the weightsjacob.goldLive Signal Community · 4h old · seen in 2 sources
- 32A Few Good Ideas in Programming Languagesprydt.xyzLive Signal Community · 10h old · seen in 2 sources HEATING ×2
- 33Show HN: Lab – Local-first, serverless browser client for LLMslabstudio.techLive Signal Community · 4h old · via Show HN
- 34js65 - Advanced 6502 assembler with patching support based on ca65 syntaxjsnesx.github.ioLive Signal Community · 4h old · via Lobsters
- 35Presentation: Investigating the WikiSwarmlesswrong.comLive Signal Community · 5h old · via LessWrong
- 36Show HN: AI-Powered Word Guessing Gamesynapse.akshayr.xyzLive Signal Community · 5h old · via Show HN
- 37Linux Zoom Client Proactively Reads X11 Clipboardhachyderm.ioLive Signal Community · 11h old · seen in 2 sources HEATING ×2
- 38Show HN: I use AirPods as wireless IMUs to track discs, bowling balls and runskinapod.comLive Signal Community · 5h old · via Show HN
- 39Show HN: Stackray – Detect any website's tech stackstackray.appLive Signal Community · 5h old · via Show HN
- 40Show HN: Ember – a strategy game where everyone moves at onceplayember.ggLive Signal Community · 5h old · via Show HN
How Radar works
The gradient. Stories travel through the AI ecosystem in a rough sequence: they appear first in papers (arXiv, Hugging Face) or repos (GitHub), then surface in community discussion (Hacker News, Lobsters, LessWrong), then get interpreted in analysis (operator newsletters and essays). Radar ranks by where a story sits on that gradient. The earlier it is caught, the more lead time you have before it reaches consensus. The spectrum at the top of the page shows the last 24 hours of that flow — every story that entered the watched feeds, counted at the station where it first appeared.
The score. Each story gets a recency-decayed signal score. Every source tier carries a weight and a half-life, so live signal moves fast and fades fast while analysis sources move slowly and stay relevant longer. A story scores higher when two or more independent source families carry it — corroboration, not popularity, which is why five feeds of the same site count once. A small bonus applies when a story has visibly climbed the gradient. Stories past their tier's maximum age expire from the board.
Heating and lead time. A story is heating when it has surfaced in two or more independent source families. When the same story is observed first at an earlier stage and later at a higher one, Radar records the gap in hours — for example, seen on arXiv fourteen hours before it reached the Hacker News front page. That gap is the reader's lead time, and it is a measured span, never an estimate.
The honesty rule. Every value on this page is an observed fact from the feed log — which source carried a story, when, and how many. Radar makes no claim about what a story means; that is the newsletter's job. When a reading does not exist yet — no stories heating, nothing upstream — the page says so rather than inventing one.
- The spectrum band
- A filled flame tick is an early, lead-time-rich stage; a hollow tick is a later stage; a hollow ring on an empty span is a station with nothing on it yet.
- Signal class
- The tier chip: Live Signal moves fast and decays fast; analysis sources stay relevant longer. The ranker weights each accordingly.
- heating
- The same story in two or more independent source families — not two feeds of one site.
- seen in / via
- How many families carried it, or the single source that did.
Once a week, the operator read on what actually changed and the move it calls for. No digest, no hype.
Read the newsletter →