[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-benchmark-exposes-cracks-in-ai-multiplayer-world-models":10,"sections":34},{"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":24,"tags":25,"sources":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},10892,"benchmark-exposes-cracks-in-ai-multiplayer-world-models","Benchmark Exposes Cracks in AI Multiplayer World Models","A new Minecraft-based benchmark finds AI-generated multiplayer worlds still fail basic persistence and structural consistency tests.","A new benchmark finds that AI-generated multiplayer worlds still cannot agree on what is actually happening to the people inside them.\n\nResearchers built MultiWorldBench, a diagnostic test set built inside Minecraft with 495 case configurations spread across seven task suites and ten capabilities, including independent control, cross-view motion, shared-state synchronization, persistence, structural reasoning, concurrent interaction, and delayed revisit. They ran three generated world-building systems, Gamma-World, Solaris, and MineWorld, through the test and compared them to a real reference engine called Engine GT. Gamma-World came out on top among the generated systems with a ten-capability average of 21.39, Solaris followed at 20.88, and MineWorld trailed far behind at 1.89. Engine GT, the actual game engine, scored 91.69.\n\nThe gap matters because every generated system topped out at 8.00 on keeping world state persistent and 1.33 on structural consistency, and none could reliably handle spatial reasoning or remember what a building was supposed to be after players looked away and came back. That is the whole premise of a shared multiplayer world: two people need to see the same thing, even after one of them leaves the room. Human judges who reviewed the same footage ranked the systems in the identical order, with a 0.96 correlation to the automated scores, so this isn't a quirky metric nobody else would agree with.\n\nIndividually, these views can look convincing. Put two players in the same generated world, though, and the illusion falls apart fast, which is a useful reality check for anyone pitching AI-built game worlds as multiplayer ready.","[\"ai\",\"world-models\",\"benchmarks\",\"minecraft\"]","2026-10-09T04:00:00.000Z","2026-10-09T20:20:58.334Z","2026-10-09T20:21:04.264Z","published",null,[],"ai",[24,26,27,28],"world-models","benchmarks","minecraft",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2610.11723",0,{"sections":35},[36,39,43,48,53,58,62,67,72,77,82,87,92,97],{"name":37,"slug":24,"count":38,"latest_published_at":18},"AI",6619,{"name":40,"slug":41,"count":42,"latest_published_at":18},"Security","security",926,{"name":44,"slug":45,"count":46,"latest_published_at":47},"Policy","policy",486,"2026-10-08T22:40:11.000Z",{"name":49,"slug":50,"count":51,"latest_published_at":52},"Deals","deals",474,"2026-10-08T22:00:00.000Z",{"name":54,"slug":55,"count":56,"latest_published_at":57},"Hardware","hardware",229,"2026-10-08T20:47:10.000Z",{"name":59,"slug":60,"count":61,"latest_published_at":18},"Science","science",192,{"name":63,"slug":64,"count":65,"latest_published_at":66},"Consumer Tech","consumer-tech",181,"2026-10-08T23:26:35.000Z",{"name":68,"slug":69,"count":70,"latest_published_at":71},"Startups","startups",117,"2026-10-08T16:45:00.000Z",{"name":73,"slug":74,"count":75,"latest_published_at":76},"Software","software",114,"2026-10-08T17:57:01.000Z",{"name":78,"slug":79,"count":80,"latest_published_at":81},"Dev Tools","dev-tools",105,"2026-10-07T16:59:11.000Z",{"name":83,"slug":84,"count":85,"latest_published_at":86},"General","general",66,"2026-10-09T04:46:11.000Z",{"name":88,"slug":89,"count":90,"latest_published_at":91},"Gaming","gaming",58,"2026-10-08T20:08:45.000Z",{"name":93,"slug":94,"count":95,"latest_published_at":96},"Reviews","reviews",34,"2026-10-08T14:00:22.000Z",{"name":98,"slug":99,"count":100,"latest_published_at":101},"How-To","how-to",8,"2026-10-05T09:00:00.000Z"]