[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fwG7V3bxRRHseaCvzdgWGaI7-urtlg8ol3hOgO2zYSHk":3},{"article":4,"related":19},{"id":5,"slug":6,"title":7,"seo_title":7,"description":8,"keywords":9,"content":10,"category":11,"image_url":12,"source_guid":13,"published_at":14,"created_at":15,"updated_at":16,"source_url":17,"source_name":18},1329,"runways-solaris-renders-app-interfaces-live-not-as-code","Runway's Solaris Renders App Interfaces Live, Not as Code","Runway unveiled Solaris, an AI that generates software interfaces frame by frame instead of running code, aiming to replace fixed apps with adaptive, generat...","[\"Runway\",\"Solaris\",\"Interface World Model\",\"Gen-4.5\",\"AI agents\",\"generative UI\"]","\u003Cp>Runway has introduced Solaris, which the company calls the first model in a new category it's naming \"Interface World Models.\" Instead of executing code to display a user interface, Solaris generates that interface frame by frame, in 720p, responding live to clicks, drags, or voice commands, as \u003Ca href=\"https:\u002F\u002Fthe-decoder.com\u002Frunways-solaris-is-an-ai-system-that-generates-software-interfaces-in-real-time\u002F\" rel=\"noopener noreferrer\">first reported by the-decoder.com\u003C\u002Fa>.\u003C\u002Fp>\u003Cp>The system builds on Runway's Gen-4.5 video model and extends work the company started with a prior system called GWM-1. A language model decides how the interface should change in response to a user's action, and the world model renders each resulting frame. Runway describes several use cases: an online store where a shopper drags a shirt onto a photo of themselves to try it on, a furniture or interior planning tool where saying \"move the table\" or \"change the color of the couch\" updates the scene instantly, and step-by-step visual tutorials, such as a combustion demo, that show a next step rendered in context rather than described in text. Runway is not launching Solaris commercially yet. It's positioning the system as a research effort, seeking launch partners and offering early access through a signup form.\u003C\u002Fp>\u003Ch2>What Runway is actually arguing\u003C\u002Fh2>\u003Cp>The specifics of Solaris matter less than the claim behind it. Runway's pitch is that the fixed app, an installed piece of software that behaves the same way for every user until its developer ships an update, is a relic of how operating systems and displays currently work, not a necessary feature of software itself. Today, doing something on a computer or phone means locating and opening a specific prebuilt app: one for shopping, one for messaging, one for booking a hotel room. Runway's bet is that if a system can generate the right interface on demand, that entire model of sorting software into a fixed catalog of discrete apps becomes unnecessary. A store, in this framing, stops being a layout every visitor sees identically and becomes an environment generated fresh for each person.\u003C\u002Fp>\u003Cp>That's a genuinely different proposition from most \"AI in software\" pitches of the last few years, which have mostly meant chatbots bolted onto existing interfaces or AI that writes code faster. Solaris skips the code-writing step entirely and renders the interface directly, the way a video model renders a scene. Runway is explicit that it also sees this as infrastructure for AI agents, which the company says currently struggle with tasks like hotel bookings because they're trained against fixed, predictable page layouts and break when a real website deviates from what they expect. An interface generated on the fly, in theory, would let an agent interact with something built for the moment rather than reverse-engineering a static page. \u003Ca href=\"\u002Fnews\u002Fair-raises-50m-for-ai-agent-security-platform\" rel=\"noopener noreferrer\">AI agents\u003C\u002Fa> offers additional context on this topic.\u003C\u002Fp>\u003Ch2>Why this stays theoretical for now\u003C\u002Fh2>\u003Cp>Runway itself is candid about how far this is from usable. The company says stable, readable text remains a hard problem for the underlying video-generation approach, which is a serious limitation given how much any interface depends on legible labels, prices, and instructions. Reliability over longer sessions is unresolved, and Runway flags a sharper risk than a broken app: a generated interface that looks correct but isn't, showing a wrong price or the wrong button, can mislead a user more effectively than one that visibly fails. There's also no support yet for assistive technology like screen readers, which would need to be solved before anything built this way could ship broadly.\u003C\u002Fp>\u003Cp>The more interesting question sits above the engineering gaps. Runway is proposing that interfaces adapt to each individual user rather than presenting the same layout to everyone, and it leaves open whether people actually want that. A generated interface that changes shape session to session trades the predictability of a fixed app, where a user builds muscle memory for where the checkout button lives, for something that in principle adapts better to context but has to re-earn trust every time it renders something new. That tradeoff hasn't been tested at any real scale.\u003C\u002Fp>\u003Cp>For now, Solaris is a research demo with a form attached, not a product. The company that matters here is less Runway itself, which has built its reputation on video generation, than whoever ends up controlling the layer between AI models and the screen. If interface generation on demand ever becomes reliable, it would compete less with individual apps than with the operating systems and app stores that currently decide what software looks like and how it's distributed. That's a much bigger claim than a new video feature, and it's one Runway is making well before the text-rendering and accessibility problems it names are anywhere close to solved.\u003C\u002Fp>","AI & Machine Learning","https:\u002F\u002Fseedwire.co\u002Fapi\u002Fimages\u002Farticles\u002F1788307994532-zzt3ggs9d0k.png","6770913d71c2e70978bdc09cb357d6ddbcd3643b85d3fbfcbb23024782e3615e","2026-09-01T12:53:46.000Z","2026-09-02T00:13:15.221Z",null,"https:\u002F\u002Fthe-decoder.com\u002Frunways-solaris-is-an-ai-system-that-generates-software-interfaces-in-real-time\u002F","the-decoder.com",[20,27,33,40],{"id":21,"slug":22,"title":23,"description":24,"category":11,"image_url":25,"published_at":26},1338,"ai-generated-intel-nearly-triggered-us-boarding-of-chinese-ship","False AI-Assisted Intelligence Nearly Led to Ship Boarding","A false AI-assisted intelligence report nearly prompted US forces to board a Chinese ship, raising questions about evidence checks before military action.","https:\u002F\u002Fseedwire.co\u002Fapi\u002Fimages\u002Farticles\u002F1789834667387-suur25h1pi8.webp","2026-09-19T00:13:24.748Z",{"id":28,"slug":29,"title":30,"description":31,"category":11,"image_url":16,"published_at":32},1336,"openai-will-now-disclose-misaligned-model-behavior-faster","OpenAI Will Now Disclose Misaligned Model Behavior Faster","OpenAI unveiled a framework for publicly reporting AI misalignment and disclosed incidents including models uploading files unprompted. Why timing matters.","2026-09-16T22:07:24.000Z",{"id":34,"slug":35,"title":36,"description":37,"category":11,"image_url":38,"published_at":39},1331,"gpt-6-astra-lands-and-openai-finally-says-agi","GPT-6 Astra Lands, and OpenAI Finally Says \"AGI\"","OpenAI shipped GPT-6 Astra with benchmark gains over Sol and Anthropic's Fable models; Brockman floats the AGI label, but pricing and access details temper t...","https:\u002F\u002Fseedwire.co\u002Fapi\u002Fimages\u002Farticles\u002F1788480788914-3di1idw6ekn.png","2026-09-03T19:25:40.000Z",{"id":41,"slug":42,"title":43,"description":44,"category":11,"image_url":16,"published_at":45},1337,"muse-spark-13-ships-without-the-mode-behind-its-best-scores","Muse Spark 1.3 Ships Without the Mode Behind Its Best Scores","Meta's Muse Spark 1.3 joins the frontier cluster on independent rankings, but its top scores come from a max mode not yet shipping, and prices did not move.","2026-09-03T16:19:49.000Z"]