[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-ai-agent-builds-a-working-rulebook-before-it-ever-wins":10,"sections":45},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":35,"tags":36,"sources":40,"feedback":44,"feedback_at":22,"cost_usd":44,"total_tokens":44},5097,"ai-agent-builds-a-working-rulebook-before-it-ever-wins","AI Agent Builds a Working Rulebook Before It Ever Wins","Twin, a test-time AI system, infers an unknown game's rules and goal from scratch, then plays ARC-AGI-3 levels more efficiently than most human testers.","A new AI agent skips the rulebook and writes its own, then beats the game anyway.\n\nResearchers built Twin, a test-time world-model inference system in which a frontier coding agent writes an executable simulation of an unfamiliar game's rules and goal, using nothing but observation and interaction, not a hand-coded model built for that specific task. It checks its own work in a sandboxed twin of the game: the harness blocks any action until the program can reproduce every past observation, and each mismatch becomes a counterexample that patches the model. Tested on ARC-AGI-3, a benchmark of grid-based games with hidden rules, Twin cleared 179 of 183 levels (97.8%). On the 179 levels it cleared, it played more efficiently than human first-timers on 158 of them (88.3%), and it figured out the goal before ever seeing a reward signal on 156 of those levels (87.2%); the rest it solved by searching for the goal directly.\n\nARC-AGI-3 scores agents from 0 to 100 on a mix of completion and efficiency, benchmarked against humans playing each game for the first time. The base model alone scores 7.8% on that scale. Bolt on a generic off-the-shelf harness and it climbs to 61.1%. Give the same base model Twin's self-written world model and the score reaches 93.3%, with the system clearing 23 of 25 games in that run. The jump from 61% to 93% is the real story: it is not more raw model capability doing the work, it is the discipline of forcing an agent to write down and test a falsifiable theory of the game before it acts.\n\nThe paper's own admission is the honest part: building a usable world model turned out to be the easy half. Guessing what the game actually wants you to do is still the hard one.","[\"ai\",\"ai-agents\",\"benchmarks\",\"arc-agi\"]","2026-08-17T04:00:00.000Z","2026-08-17T10:28:33.416Z","2026-08-17T10:28:45.253Z","published",null,[24,30],{"id":25,"reviewer":26,"round":27,"reason":28,"status":29},"editor-r1","editor",1,"The dek attributes the 97.8% figure (level-completion rate) to being 'faster than human first-timers,' but the body shows that comparison is actually 88.3% (158\u002F179 levels); fix the dek so the stat it cites matches the claim it makes.","resolved",{"id":31,"reviewer":32,"round":33,"reason":34,"status":29},"publisher-r2","publisher",2,"Reported percentages don't match the stated fractions (158\u002F183 is 86.3% not 88.3%, 156\u002F183 is 85.2% not 87.2%, and 23\u002F25 is 92% not 93.3%), indicating unresolved numeric errors in the body.","ai",[35,37,38,39],"ai-agents","benchmarks","arc-agi",[41],{"name":42,"url":43},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2608.14490",0,{"sections":46},[47,51,55,60,65,70,75,80,85,90,95,100,105,110],{"name":48,"slug":35,"count":49,"latest_published_at":50},"AI",3293,"2026-08-20T04:00:00.000Z",{"name":52,"slug":53,"count":54,"latest_published_at":50},"Security","security",435,{"name":56,"slug":57,"count":58,"latest_published_at":59},"Policy","policy",210,"2026-08-19T09:32:27.000Z",{"name":61,"slug":62,"count":63,"latest_published_at":64},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":66,"slug":67,"count":68,"latest_published_at":69},"Hardware","hardware",140,"2026-08-19T18:25:42.000Z",{"name":71,"slug":72,"count":73,"latest_published_at":74},"Consumer Tech","consumer-tech",95,"2026-08-18T16:05:00.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":79},"Science","science",90,"2026-08-19T18:41:02.000Z",{"name":81,"slug":82,"count":83,"latest_published_at":84},"Software","software",73,"2026-08-18T07:51:50.000Z",{"name":86,"slug":87,"count":88,"latest_published_at":89},"Dev Tools","dev-tools",69,"2026-08-18T04:00:00.000Z",{"name":91,"slug":92,"count":93,"latest_published_at":94},"Startups","startups",47,"2026-08-19T19:13:46.000Z",{"name":96,"slug":97,"count":98,"latest_published_at":99},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":101,"slug":102,"count":103,"latest_published_at":104},"General","general",33,"2026-08-18T22:18:13.000Z",{"name":106,"slug":107,"count":108,"latest_published_at":109},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":111,"slug":112,"count":113,"latest_published_at":114},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]