[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-why-your-ai-agent-does-something-different-every-time":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},3580,"why-your-ai-agent-does-something-different-every-time","Why Your AI Agent Does Something Different Every Time","A new paper breaks down the layered sources of randomness in agentic AI systems, from token sampling to live data and noisy infrastructure.","AI agents are not deterministic machines, and a new tutorial paper explains exactly why.\n\nResearchers publishing on arXiv argue that the variability people observe in agentic AI systems comes from multiple distinct layers that developers routinely conflate. At the foundation is token sampling: a model converts internal scores into probabilities, then draws from those probabilities using a pseudo-random number generator. A single different token early in a response can cascade into a different tool call, a different code path, or an entirely different plan. But that is just the start. On top of sampling, agents also vary because of changing environments, live data feeds, serving infrastructure quirks, batch processing effects, and low-level numerical differences across hardware.\n\nThe distinction matters because developers trying to debug or reproduce an agent failure often blame randomness when the real culprit is a shifting external environment, or vice versa. The paper frames reproducibility as conditional: matched conditions can tame token-sampling variance, but deployed agents face a stack of extrinsic noise sources that no temperature setting can fix. For teams building reliable agentic pipelines, that is a useful diagnostic frame, not just an academic one.\n\nThe field has spent considerable energy debating whether to run models at temperature zero for consistency - a tactic that addresses only the innermost layer of the problem. Any team that stops there and calls the system deterministic is, the paper implies, fooling itself.","[\"ai\",\"agents\",\"machine-learning\",\"research\"]","2026-07-03T04:00:00.000Z","2026-07-03T09:02:00.841Z","2026-07-03T09:02:03.750Z","published",null,[],"ai",[24,26,27,28],"agents","machine-learning","research",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2606.08998",0,{"sections":35},[36,40,45,50,55,60,65,70,75,80,85,89,94,99],{"name":37,"slug":24,"count":38,"latest_published_at":39},"AI",2590,"2026-07-16T04:00:00.000Z",{"name":41,"slug":42,"count":43,"latest_published_at":44},"Security","security",294,"2026-07-15T19:59:48.000Z",{"name":46,"slug":47,"count":48,"latest_published_at":49},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":51,"slug":52,"count":53,"latest_published_at":54},"Policy","policy",158,"2026-07-16T00:02:48.000Z",{"name":56,"slug":57,"count":58,"latest_published_at":59},"Hardware","hardware",122,"2026-07-14T19:46:26.000Z",{"name":61,"slug":62,"count":63,"latest_published_at":64},"Consumer Tech","consumer-tech",93,"2026-07-13T13:20:48.000Z",{"name":66,"slug":67,"count":68,"latest_published_at":69},"Software","software",70,"2026-07-13T19:52:25.000Z",{"name":71,"slug":72,"count":73,"latest_published_at":74},"Science","science",66,"2026-07-10T10:29:37.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":79},"Dev Tools","dev-tools",59,"2026-07-07T04:00:00.000Z",{"name":81,"slug":82,"count":83,"latest_published_at":84},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":86,"slug":87,"count":83,"latest_published_at":88},"Startups","startups","2026-06-29T20:55:50.000Z",{"name":90,"slug":91,"count":92,"latest_published_at":93},"General","general",29,"2026-07-10T22:28:58.000Z",{"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"]