--- title: New AI Just Reinvented Minecraft Worlds | daily.dev description: A solo researcher published a paper at SIGGRAPH introducing a novel infinite terrain generation technique that combines the speed of noise-based methods with... canonical: https://daily.dev/posts/new-ai-just-reinvented-minecraft-worlds-fdfszvpku twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: New AI Just Reinvented Minecraft Worlds | daily.dev og:description: A solo researcher published a paper at SIGGRAPH introducing a novel infinite terrain generation technique that combines the speed of noise-based methods with... og:url: https://daily.dev/posts/new-ai-just-reinvented-minecraft-worlds-fdfszvpku og:image: https://api.daily.dev/og/posts/FDFSZvPKU.png og:image:alt: New AI Just Reinvented Minecraft Worlds og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # New AI Just Reinvented Minecraft Worlds **[Two Minute Papers](https://daily.dev/sources/twoninutepapers)** · 6 min read · 5 upvotes · 0 comments ## Summary A solo researcher published a paper at SIGGRAPH introducing a novel infinite terrain generation technique that combines the speed of noise-based methods with the learning capability of AI-based approaches. The system uses diffusion models with a windowed neighbor-averaging formula so generation cost stays constant regardless of world size. A second innovation called Laplacian re-extraction denoising handles multi-scale height variation, allowing both mountain ranges and fine details like river banks to coexist. The model was trained in two weeks on a consumer GPU, and both the code and a Minecraft mod are freely available. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How does this new AI terrain generation technique achieve infinite scale without slowing down? It decouples the cost of generating a new region from the total size of the world by only querying overlapping neighboring windows rather than the entire world state. Each new region is generated by running diffusion denoising on nearby windows and taking their weighted average, so query cost stays constant even as the generated world grows arbitrarily large, enabling instant teleportation across millions of miles. _daily.dev surfaces deep technical breakdowns like this for developers tracking generative AI techniques in graphics._ ### How does this terrain generator handle both huge mountains and small details like riverbanks at the same time? It uses a Laplacian-style trick that separates terrain into multiple scales, generating a large-scale image for big features like mountains and a separate fine-scale pass for small details like ridges and river banks, then combining both. Standard diffusion denoising on raw height values fails at this because it can only focus on one scale at a time. _developers evaluating diffusion techniques for graphics can follow breakdowns like this on daily.dev._ ### What hardware and training time did this infinite terrain generation AI model require? The model was trained in two weeks and ran interactively on a four-year-old consumer GPU. It was developed by a single independent scientist and published at SIGGRAPH, the leading computer graphics conference, with the code and a Minecraft mod released for free. _anyone tracking efficient generative AI research can follow updates like this via daily.dev._ ## Similar posts on daily.dev - [\[2512.08309\] Terrain Diffusion: A Diffusion-Based Successor to Perlin Noise in Infinite, Real-Time Terrain Generation](https://daily.dev/posts/2512-08309-terrain-diffusion-a-diffusion-based-successor-to-perlin-noise-in-infinite-real-time-t-psh7s6i4o) · Hacker News · 2 upvotes · 0 comments - [Towards Data Science](https://daily.dev/posts/towards-data-science-n9ra6ei3l) · Towards Data Science · 0 upvotes · 0 comments - [World Models: The Quiet Race to Build AI That Understands Reality, Not Just Language](https://daily.dev/posts/world-models-the-quiet-race-to-build-ai-that-understands-reality-not-just-language-2pfjhmqhl) · Medium · 0 upvotes · 0 comments --- Tags: [#game-development](https://daily.dev/tags/game-development), [#graphics-programming](https://daily.dev/tags/graphics-programming), [#diffusion-models](https://daily.dev/tags/diffusion-models) [View this post on daily.dev](https://daily.dev/posts/new-ai-just-reinvented-minecraft-worlds-fdfszvpku) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"New AI Just Reinvented Minecraft Worlds","url":"https://daily.dev/posts/new-ai-just-reinvented-minecraft-worlds-fdfszvpku","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/new-ai-just-reinvented-minecraft-worlds-fdfszvpku"},"datePublished":"2026-07-12T15:55:27.056Z","dateModified":"2026-09-14T06:13:02.169Z","description":"A solo researcher published a paper at SIGGRAPH introducing a novel infinite terrain generation technique that combines the speed of noise-based methods with...","image":"https://i.ytimg.com/vi/Ae9q7KsRbuI/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/Ae9q7KsRbuI/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Two Minute Papers","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Two Minute Papers","logo":"https://media.daily.dev/image/upload/s--8KXsSJ5q--/f_auto/v1711188923/logos/twoninutepapers","url":"https://daily.dev/sources/twoninutepapers"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/new-ai-just-reinvented-minecraft-worlds-fdfszvpku","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":5},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"game-development,graphics-programming,diffusion-models","timeRequired":"PT6M","video":{"@type":"VideoObject","name":"New AI Just Reinvented Minecraft Worlds","description":"A solo researcher published a paper at SIGGRAPH introducing a novel infinite terrain generation technique that combines the speed of noise-based methods with...","thumbnailUrl":"https://i.ytimg.com/vi/Ae9q7KsRbuI/sddefault.jpg","uploadDate":"2026-07-12T15:55:27.056Z","duration":"PT6M","url":"https://api.daily.dev/r/FDFSZvPKU","embedUrl":"https://www.youtube.com/embed/Ae9q7KsRbuI"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Two Minute Papers","item":"https://daily.dev/sources/twoninutepapers"},{"@type":"ListItem","position":3,"name":"New AI Just Reinvented Minecraft Worlds"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/new-ai-just-reinvented-minecraft-worlds-fdfszvpku#faq","mainEntity":[{"@type":"Question","name":"How does this new AI terrain generation technique achieve infinite scale without slowing down?","acceptedAnswer":{"@type":"Answer","text":"It decouples the cost of generating a new region from the total size of the world by only querying overlapping neighboring windows rather than the entire world state. Each new region is generated by running diffusion denoising on nearby windows and taking their weighted average, so query cost stays constant even as the generated world grows arbitrarily large, enabling instant teleportation across millions of miles. daily.dev surfaces deep technical breakdowns like this for developers tracking generative AI techniques in graphics."}},{"@type":"Question","name":"How does this terrain generator handle both huge mountains and small details like riverbanks at the same time?","acceptedAnswer":{"@type":"Answer","text":"It uses a Laplacian-style trick that separates terrain into multiple scales, generating a large-scale image for big features like mountains and a separate fine-scale pass for small details like ridges and river banks, then combining both. Standard diffusion denoising on raw height values fails at this because it can only focus on one scale at a time. developers evaluating diffusion techniques for graphics can follow breakdowns like this on daily.dev."}},{"@type":"Question","name":"What hardware and training time did this infinite terrain generation AI model require?","acceptedAnswer":{"@type":"Answer","text":"The model was trained in two weeks and ran interactively on a four-year-old consumer GPU. It was developed by a single independent scientist and published at SIGGRAPH, the leading computer graphics conference, with the code and a Minecraft mod released for free. anyone tracking efficient generative AI research can follow updates like this via daily.dev."}}]} ```