What's moving through AI, ranked by lead time.
Radar is an hourly 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. 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 23 Aug 2026, 23:38 UTC · 66/78 feeds live · 18,633 tracked since 12 Jun
- 1 Declarative WebGPU with S-expressions hugodaniel.com · 3h
- 2 Everything I own, owned schlarp.com · 1h
- 3 In search of natural features lesswrong.com · 1h
- 4 The Planet Now Has More Trees Than It Did 35 Years Ago (2018) psmag.com · 2h
- 5 A complex structure on S^6 [pdf] alpo.ge · 2h
172 stories entered the watched feeds in the last 24 hours — 10 in papers, 8 in repos, 143 in community, 11 in analysis. The scored board skews to Community (36 of 40). 3 stories are heating across independent sources, and the median lead across the ledger's 20 observed fires is 5h — measured, not estimated. Freshest receipt, 5d ago: caught in GitHub Trending Python 25d before Show HN carried it. The oldest survivor has held the board for 5d.
- Repos → Community +25d lead GitHub Trending Python first · then Show HN 5d agomoorcheh-ai/memanto
- Repos → Community +32d lead GitHub Trending Python first · then Hacker News front 5d agoRyanCodrai/turbovec
- Community → Analysis +4h lead Hacker News front first · then Stratechery 12d agoNvidia's Risky Business
- 1Declarative WebGPU with S-expressionshugodaniel.comLive Signal Community · 3h old · seen in 2 sources HEATING ×2
- 2
- 3
- 4The Planet Now Has More Trees Than It Did 35 Years Ago (2018)psmag.comLive Signal Community · 2h old · via Hacker News front
- 5
- 6Implementation of GPT-2 in pure CMakegithub.comLive Signal Community · 2h old · via Hacker News front
- 7
- 8The text mode lie: why modern TUIs are a nightmare for accessibilityosnews.comLive Signal Community · 3h old · via Lobsters
- 9I spent $266 and four AI models to own my tablet. GLM-5.3 finished it in a dayericpardee.github.ioLive Signal Community · 9h old · seen in 3 sources HEATING ×2
- 10Decoding silent reading from non-invasive EEGarxiv.orgLive Signal Community · 3h old · via Hacker News front
- 11I like 'em thickexperimental-history.comContrarian Analysis · 5d old · seen in 2 sources HEATING ×2
- 12
- 13Google Workspace thinks my domain is an email providerblog.elis.ccLive Signal Community · 4h old · via Hacker News front
- 14Replicating Reddit's best feature on other forumsxavd.idLive Signal Community · 4h old · via Lobsters
- 15My experience as a first time speakerblog.guillheu.devLive Signal Community · 4h old · via Lobsters
- 16Fable and the End of the Free Lunchdbreunig.comLive Signal Community · 5h old · via Hacker News front
- 17Wild AI-related reliability incidents are comingsurfingcomplexity.blogLive Signal Community · 5h old · via Lobsters
- 18Over 170k Nonprofits Lost All Their Data. Is Microsoft to Blame?slate.comLive Signal Community · 5h old · via Hacker News front
- 19
- 20
- 21Etched Sohu vs. Nvidia: Transformer ASIC vs. GPU (2026) – Spheron Blogspheron.networkLive Signal Community · 5h old · via Hacker News front
- 22Anthropic's best AI model struggles to attract users as cheaper tools thriveft.comLive Signal Community · 5h old · via Hacker News front
- 23AI and Infrastructure Engineeringomegion.devLive Signal Community · 6h old · via Hacker News front
- 24
- 25Predicting AI model release dates with statsreleaseoracle.xyzLive Signal Community · 6h old · via Hacker News front
- 26I turned Unix talk from 1983 into the interface for my AIen.andros.devLive Signal Community · 6h old · via Hacker News front
- 27My agent.md to improve LLM-assisted code qualityfabiensanglard.netLive Signal Community · 6h old · via Hacker News front
- 28Anthropic’s best AI model struggles to attract users as cheaper tools thrivesimonwillison.netAI Firehose Analysis · 3h old · via Simon Willison
- 29
- 30We must not grant AI agents legal personhoodft.comLive Signal Community · 6h old · via Hacker News front
- 31The Remote Work Challenge: Lessons from 5 Citiespew.orgLive Signal Community · 6h old · via Hacker News front
- 32Erik Brynjolfsson says an AI "job apocalypse" is unlikelywpintelligence.washingtonpost.comLive Signal Community · 7h old · via Hacker News front
- 33A website for debloated open source alternativesdebloat.devLive Signal Community · 7h old · via Hacker News front
- 34
- 35
- 36Phoenix tried a reflective coating on black asphalt; noon surface heat fell 12°Feconomictimes.indiatimes.comLive Signal Community · 7h old · via Hacker News front
- 37PSA: There's a third option in the "measure problem"lesswrong.comLive Signal Community · 7h old · via LessWrong
- 38
- 39Utilities as Legendre duals of probabilitieslesswrong.comLive Signal Community · 7h old · via LessWrong
- 40Steve Jobs and hiring | Anthropic cofounder's cold emailtechemails.comPrimary Analysis Analysis · 31h old · via Internal Tech Emails
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 →