[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-new-robot-ai-framework-lets-skills-reuse-instead-of-retrain":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},5693,"new-robot-ai-framework-lets-skills-reuse-instead-of-retrain","New Robot AI Framework Lets Skills Reuse Instead of Retrain","A new architecture called Teach-and-Grow stores robot skills as reusable blocks, aiming to cut the retraining every new task normally demands.","A new robot-learning framework skips the usual retrain-from-scratch routine when a machine hits something unfamiliar.\n\nResearchers describe Teach-and-Grow Learning (TGL), an agent-centered architecture that turns a handful of successful demonstrations into reusable Skill Blocks, closed-loop behaviors tied to specific subgoals. A multimodal agent stores these blocks in a Skill Library, then pulls from a separate Experience Memory that tracks what worked, what failed, and how problems were fixed. When the robot lands in a new scene, the agent grounds and combines existing blocks, picks the right tools, watches what actually happens, and adjusts course if execution drifts from intent. On the LIBERO benchmark, the system posted state-of-the-art results, and separate tests showed it could induce new skills, reuse old ones, and adapt without retraining a task-specific policy.\n\nThat last part is the real pitch. End-to-end vision-language-action models are only as reliable as their training coverage, and every gap, whether a new object, sensor, or contact case, has traditionally meant collecting more physical robot data and retraining, what the authors call the retraining tax. Because embodied data requires actually operating machines, that tax runs far more expensive than fine-tuning a language model on more text.\n\nThe paper also floats a scaling-law hypothesis: more reusable experience should push future errors and teaching effort down predictably. That is worth watching, not banking on. It is demonstrated on a simulated benchmark, not a warehouse floor, and predictable scaling has a way of getting messier once real robots and real clutter get involved.","[\"robotics\",\"ai\",\"machine-learning\",\"vla-models\"]","2026-08-19T04:00:00.000Z","2026-08-19T11:51:03.986Z","2026-08-19T11:51:15.879Z","published",null,[],"ai",[26,24,27,28],"robotics","machine-learning","vla-models",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2608.17209",0,{"sections":35},[36,40,44,49,54,59,64,69,74,79,84,89,94,99],{"name":37,"slug":24,"count":38,"latest_published_at":39},"AI",3293,"2026-08-20T04:00:00.000Z",{"name":41,"slug":42,"count":43,"latest_published_at":39},"Security","security",435,{"name":45,"slug":46,"count":47,"latest_published_at":48},"Policy","policy",210,"2026-08-19T09:32:27.000Z",{"name":50,"slug":51,"count":52,"latest_published_at":53},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":55,"slug":56,"count":57,"latest_published_at":58},"Hardware","hardware",140,"2026-08-19T18:25:42.000Z",{"name":60,"slug":61,"count":62,"latest_published_at":63},"Consumer Tech","consumer-tech",95,"2026-08-18T16:05:00.000Z",{"name":65,"slug":66,"count":67,"latest_published_at":68},"Science","science",90,"2026-08-19T18:41:02.000Z",{"name":70,"slug":71,"count":72,"latest_published_at":73},"Software","software",73,"2026-08-18T07:51:50.000Z",{"name":75,"slug":76,"count":77,"latest_published_at":78},"Dev Tools","dev-tools",69,"2026-08-18T04:00:00.000Z",{"name":80,"slug":81,"count":82,"latest_published_at":83},"Startups","startups",47,"2026-08-19T19:13:46.000Z",{"name":85,"slug":86,"count":87,"latest_published_at":88},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":90,"slug":91,"count":92,"latest_published_at":93},"General","general",33,"2026-08-18T22:18:13.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":100,"slug":101,"count":102,"latest_published_at":103},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]