source&pool
A daily wire of long-form journalism, video, and discourse — filed, tagged, and laid out flat.
VOL. I·NO. 01
THURSDAY, SEPTEMBER 24, 2026
Hacker News3454X 主题热门3361CNBC59YahooFinance57aihot539to5Mac46MacRumors46IGN38Verge37Kotaku32TechCrunch25AndroidAuthority23Gematsu19NintendoLife19Engadget179to5Google15BusinessInsider15WarhammerCommunity15Eurogamer14Guardian14USAToday14ArsTechnica13Fortune11FoxBusiness11Polygon11PushSquare11Wccftech11Investor'sBusinessDaily10NPR10ABC9Gizmodo9NBC9Notebookcheck9SeekingAlpha9TechPowerUp9CBS8CNN8MotleyFool8GSMArena8NintendoEverything8VideoCardz8VideoGamesChronicle8AppleInsider7CNET7CoinDesk7Fox7PureXbox7Yahoo7BleepingComputer6Mashable6NewYorkPost6Conversation6AlJazeera5bgr5TechSpot5Tom'sGuide5WIRED5Aftermath4AndroidPolice4Deadline4Motor14XBOXWire4PetaPixel4PlayStationLifeStyle4Pokemon4SamMobile4SouthChinaMorningPost4Variety4WSB-TV4BellofLostSouls3DroidLife3GamesIndustry.biz3Hackaday3HollywoodReporter3HuffPost3Nature3PokeBeach3RockPaperShotgun3RPGSite3SeattleTimes3SFGATE3SlashGear3WhatHi-Fi?3404Media26abcPhiladelphia280Level2ABC7LosAngeles2AndroidCentral2AZFamily2Benzinga2Currently2DCRainmaker2DigitalFoundry2EventHubs2Futurism2GameInformer2HouseDigest2InsiderGaming2Jalopnik2LosAngelesTimes2Lifehacker2PCMag2qz2SimsCommunity2Slate2TimeExtension2TODAY2TweakTown2YahooFinanceUK2UploadVR2WindowsCentral224/7WallSt.1BusinessInsiderAfrica1Alternet1AndroidHeadlines1ArizonaSports1BikeRadar1BloodyDisgusting1BostonGlobe1BusinessTimes1Yahoo!FinanceCanada1CalMatters1CarBuzz1cbn1ChromeUnboxed1Chron1ClaimDepot1ColoradoSun1Skin.ClubCommunity1ChristianScienceMonitor1CyberSecurityNews1Cyclingnews1DailyDownforce1DaringFireball1DarkHorizons1Decrypt1Defense1denver71Designboom1DigitalCameraWorld1DirtonDirt1DSOGaming1Electrek1empireonline1GameGPU1erictopol.substack1Fangoria1FOX191DetroitFreePress1GameRant1GameWorldObserver1GAMINGbible1GamingOnLinux1AAAGasPrices1GearPatrol1GeekWire1GeekyGadgets1Global1Gothamist1Hackster.io1Hodinkee1HoustonChronicle1Independent1InterestingEngineering1Invezz1MacObserver1Magic:Gathering1MakeUseOf1Mashed1Mercury1MLive1MorningBrew1Motorsport1MP1st1MyNintendo1BloombergLaw1SemiAnalysis1Newsshooter1Newsweek1NintendoWire1nrn1CrudeOilPricesToday1OneMileataTime1OregonPublicBroadcasting1OregonLive1PageSix1PersonaCentral1politico.eu1QuantaMagazine1Realtor1Road&Track1RockstarINTEL1Salon1SeattleRed1Semafor1SanFranciscoChronicle1SimpleFlying1GhostHowls1SlippedDisc1SoraNews241SpaceNews1YahooTech1the5krunner1DailyBeast1DailyMeal1Drive1Hacker1Hindu1Intercept1Register1Times1TimesofIndia1TimesUnion1TMZ1TopGear1PCMagUK1VisualCapitalist1Vulture1WCVB1WFMZ1WHYY1WKYT1YGOrganization1YourTango1
  1. 001Hacker NewsSEP · 22English

    Koòrdinate Thinking

    Koòrdinate Thinking is a method delivered as self-contained Markdown files (available in Italian and English) that can be used as system prompts, attachments, or documents without external dependencies. The approach emphasizes keeping all context within the conversation, with no lazy loading or file system requirements, and includes operational sections plus rationale and guidance for adaptation to different platforms.

    By UnmarkedPM
  2. 002Hacker NewsSEP · 22English

    Same prompt, same endpoint, different score

    A study tested whether a small prompt edit improved tool-calling accuracy on 100 test cases from BFCL V4, a UC Berkeley benchmark. While the revised prompt successfully reduced invented optional arguments, the overall score gain was too small to be considered significant, and results remained inconsistent across runs even with fixed settings and providers.

    By Khuyen Tran
  3. 003Hacker NewsSEP · 21English

    You could have built Jev

    Jev is essentially a Large Language Model designed to return a single token, functioning as a specialized classifier for tasks like determining whether a food item is a sandwich. The article explains the technical mechanics of how such a system works, breaking down LLM token generation and demonstrating how one could build a similar project using existing open-source models and prompt engineering techniques.

    By andsoitis
  4. 004Hacker NewsSEP · 19English

    You could have built Jev

    Jev is a Large Language Model that generates a single token to answer classification questions, comparing user-provided data against predefined criteria. The article explains the technical mechanics of how such a system works and discusses similar projects that have emerged.

    By baobabKoodaa
  5. 005Hacker NewsSEP · 19English

    There is no model called taste-1

    Command Code markets a feature called "taste-1" as a proprietary neuro-symbolic AI model, but investigation reveals it does not exist as a distinct model. Taste is actually a prompt constructed from user git history and conversation context, sent to whatever underlying model the user already has selected. The feature works by analyzing code diffs, extracting style patterns, and injecting guidelines into the system prompt.

    By infomiho
  6. 006Hacker NewsSEP · 18English

    Alea iacta non est: rerolling the LLM dice

    A developer created a tool called 'reroll' that generates LLM responses five times to test consistency and analyze variation patterns. The tool uses Haiku to categorize differences—whether answers are uniform, mostly uniform, split, or highly divergent—and identifies whether variations stem from different interpretations, different next moves, or contradictions that suggest hallucinations.

    By Roland Szabo
  7. 007Hacker NewsSEP · 18English

    BuiltByVibe – Directory of vibe-coded apps with a free MCP server

    BuiltByVibe is a curated directory of AI-generated software projects where users can share and discover applications created through prompt engineering. The platform hosts 56 projects with community engagement features and includes a free MCP server.

    By notevenbuttery
  8. 008Hacker NewsSEP · 17English

    Classic ML to Cope with Dumb LLM Judges (2025)

    A developer uses local LLMs to evaluate e-commerce search relevance by comparing product attributes against human-labeled data from Wayfair's WANDS dataset. They test various prompting strategies—including forced decisions, consistency checking, and single-attribute analysis—to create a reliable laptop-based search judge that can guide relevance improvements without expensive API costs.

    By Doug Turnbull