<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: WEI LIU</title>
    <description>The latest articles on DEV Community by WEI LIU (@wei_liu_33dd3614a31eda120).</description>
    <link>https://dev.to/wei_liu_33dd3614a31eda120</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4040967%2F4828526e-a319-46c3-9b72-f6e19cc4f5a6.jpg</url>
      <title>DEV Community: WEI LIU</title>
      <link>https://dev.to/wei_liu_33dd3614a31eda120</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/wei_liu_33dd3614a31eda120"/>
    <language>en</language>
    <item>
      <title>HackerTube AI: A Human-Curated Video Archive for Curious Learners</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Wed, 04 Nov 2026 20:53:04 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/hackertube-ai-a-human-curated-video-archive-for-curious-learners-4cb3</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/hackertube-ai-a-human-curated-video-archive-for-curious-learners-4cb3</guid>
      <description>&lt;p&gt;HackerTube AI is a community-curated video archive inspired by Hacker News. It helps learners, developers, researchers, and educators discover thoughtful explanations, lectures, tutorials, documentaries, and conference talks.&lt;/p&gt;

&lt;p&gt;Instead of relying entirely on automated recommendations, HackerTube AI highlights videos that have attracted attention and discussion from a technically curious community. You can search by title, creator, or subject, then narrow results by broad fields and specific topics.&lt;/p&gt;

&lt;p&gt;The platform also includes structured learning paths, such as routes from mathematical foundations and neural networks toward deep learning and language models. Its Curiosity Atlas helps connect ideas across disciplines.&lt;/p&gt;

&lt;p&gt;Explore HackerTube AI: &lt;a href="https://hacker-tube.com" rel="noopener noreferrer"&gt;https://hacker-tube.com&lt;/a&gt;&lt;/p&gt;

</description>
      <category>ai</category>
      <category>webdev</category>
      <category>programming</category>
      <category>learning</category>
    </item>
    <item>
      <title>A Practical Prompt Workflow for Cinematic Photo and Video Concepts</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Wed, 04 Nov 2026 13:05:05 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/a-practical-prompt-workflow-for-cinematic-photo-and-video-concepts-1oaf</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/a-practical-prompt-workflow-for-cinematic-photo-and-video-concepts-1oaf</guid>
      <description>&lt;p&gt;Creative direction gets easier when a prompt reads like a small production brief instead of a pile of adjectives. Whether you are exploring a portrait, a fashion edit, a festival greeting, or a short video idea, the same sequence helps: define the subject, describe the scene, set the visual mood, and then add one controlled movement or camera cue.&lt;/p&gt;

&lt;h2&gt;
  
  
  Start with the subject and setting
&lt;/h2&gt;

&lt;p&gt;Begin with what must stay consistent. Name the subject, the environment, the time of day, and the intended point of attention. For example, a cinematic portrait brief might specify a person in an emerald saree, standing near a warm window at sunset, with the face and fabric as the visual focus. Concrete nouns give an image tool a clearer starting point than a long list of vague style words.&lt;/p&gt;

&lt;h2&gt;
  
  
  Add wardrobe, light, and composition
&lt;/h2&gt;

&lt;p&gt;Next, describe the details that shape the frame: materials, colors, lighting direction, lens feel, and composition. A short list is usually easier to refine than one dense sentence. If the first result is close, change one variable—such as the light temperature or camera angle—rather than rewriting the entire prompt. That makes each iteration easier to compare.&lt;/p&gt;

&lt;h2&gt;
  
  
  Treat video as a small motion study
&lt;/h2&gt;

&lt;p&gt;For a five-second concept, choose one gentle movement. A slow camera push, drifting fabric, a small turn toward the light, or rain crossing a city scene can communicate more clearly than several simultaneous actions. Keep the motion compatible with the subject and setting, then describe the mood you want the movement to preserve.&lt;/p&gt;

&lt;h2&gt;
  
  
  Use examples as planning tools
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://vikas-editor.net/" rel="noopener noreferrer"&gt;Vikas Editor AI&lt;/a&gt; is a planning and inspiration workspace for cinematic portraits, fashion edits, festival greetings, couple photography, and photo-to-video concepts. Its templates pair a visual direction with an adaptable prompt, so creators can inspect the structure, customise an idea, and take the prepared prompt to a compatible AI tool. The site currently focuses on planning; generation is not available directly on the website.&lt;/p&gt;

&lt;h2&gt;
  
  
  A repeatable checklist
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;Subject and visual identity&lt;/li&gt;
&lt;li&gt;Location, time, and atmosphere&lt;/li&gt;
&lt;li&gt;Wardrobe, materials, and color palette&lt;/li&gt;
&lt;li&gt;Lighting direction and camera composition&lt;/li&gt;
&lt;li&gt;Mood plus one clear motion cue for video&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;This workflow keeps creative experiments focused without removing room for style. Start with a specific brief, change one detail at a time, and keep the version that best communicates the story you want the image or clip to tell.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>promptengineering</category>
      <category>tooling</category>
    </item>
    <item>
      <title>PavoAI: One Idea, Every Model, One Finished Video</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Tue, 03 Nov 2026 14:11:26 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/pavoai-one-idea-every-model-one-finished-video-335p</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/pavoai-one-idea-every-model-one-finished-video-335p</guid>
      <description>&lt;p&gt;Turning a rough idea into a finished video usually means juggling several tools, models, and editing steps. PavoAI brings that workflow into one AI video creation platform, from the first prompt to a publish-ready result.&lt;/p&gt;

&lt;p&gt;PavoAI supports text-to-video, image-to-video, AI editing, image generation, talking avatars, music, voice, subtitles, restyling, and upscaling. Creators can compare leading models such as Veo, Kling, Sora, Hailuo, Wan, Seedance, Runway, Luma, Pika, Vidu, and PixVerse from one workspace.&lt;/p&gt;

&lt;p&gt;The platform is designed for short films, product ads, UGC campaigns, music videos, anime projects, social reels, talking avatars, and cinematic trailers. Image-to-video tools can animate product photos or portraits, while reference locking helps keep characters and products consistent across shots.&lt;/p&gt;

&lt;p&gt;A scene-based timeline editor helps arrange generated clips, add subtitles and dubbing, create music, and prepare vertical or widescreen versions. PavoAI also offers starter credits without requiring a payment card, with credits that do not expire, commercial use, and watermark-free exports.&lt;/p&gt;

&lt;p&gt;Learn more about PavoAI: &lt;a href="https://pavoai.net/" rel="noopener noreferrer"&gt;https://pavoai.net/&lt;/a&gt;&lt;/p&gt;

</description>
      <category>tools</category>
    </item>
    <item>
      <title>Filmcraft: A Free, Open-Source Video Editor Written in Rust</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Tue, 27 Oct 2026 16:17:49 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/filmcraft-a-free-open-source-video-editor-written-in-rust-37c4</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/filmcraft-a-free-open-source-video-editor-written-in-rust-37c4</guid>
      <description>&lt;p&gt;Filmcraft is a free, open-source, non-linear video editor created by the ArtCraft team. It is written in Rust and runs natively on macOS, Windows, Linux, and FreeBSD. A WebAssembly version also enables browser-based editing without installation.&lt;/p&gt;

&lt;h2&gt;
  
  
  A familiar editing model
&lt;/h2&gt;

&lt;p&gt;Filmcraft is designed as an accessible alternative to Adobe Premiere Pro for creators, learners, and personal projects. The editor includes Project panels, bins, Source and Program monitors, multitrack timelines, three-point editing, and ripple, roll, slip, and slide trimming. Keyboard shortcuts follow conventions familiar to Premiere users.&lt;/p&gt;

&lt;p&gt;The project supports frame-accurate editing through an integer-based time system, including 23.976 and 29.97 drop-frame timelines. Native codec support includes H.264, HEVC, ProRes, VP9, AV1, AAC, and Opus. Export options include H.264 MP4, ProRes 422 HQ, PNG sequences, and GIF.&lt;/p&gt;

&lt;h2&gt;
  
  
  Creative tools and automation
&lt;/h2&gt;

&lt;p&gt;Filmcraft includes more than 90 video effects and approximately 30 transitions, along with GPU-composited keyframes, curves, color wheels, HSL secondaries, LUTs, HDR and log workflows, waveform and vectorscope displays, loudness meters, audio automation, and native DSP effects. Workflows can also be scripted through the CLI, JSON channel, or MCP server.&lt;/p&gt;

&lt;p&gt;For browser-based editing, users can keep their media on their own computers while working through the WebAssembly version. The project is alpha-stage, so direct .prproj import and third-party plugin support are not available yet, and very large footage can present performance concerns.&lt;/p&gt;

&lt;p&gt;If you are exploring a subscription-free editing workflow, you can learn more at &lt;a href="https://getfilmcraft.org/" rel="noopener noreferrer"&gt;getfilmcraft.org&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>video</category>
    </item>
    <item>
      <title>DeckCraft: Open-Source Presentation Software in Rust</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Mon, 26 Oct 2026 11:39:53 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/deckcraft-open-source-presentation-software-in-rust-4d88</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/deckcraft-open-source-presentation-software-in-rust-4d88</guid>
      <description>&lt;p&gt;DeckCraft is a free, open-source presentation editor for people who want a capable slide workflow without a subscription or account requirement. It is written in pure Rust and runs as both a browser-based editor and a native desktop application for macOS, Windows, and Linux.&lt;/p&gt;

&lt;p&gt;The online editor can start with a blank presentation or a sample deck. Its workspace includes slide thumbnails, a central canvas, and familiar formatting tools for people coming from PowerPoint, Keynote, or Google Slides. Decks can include text, images, charts, vector shapes, themes, slide masters, reusable layouts, sections, speaker notes, transitions, and entrance, emphasis, and exit animations.&lt;/p&gt;

&lt;p&gt;DeckCraft can open and save PPTX files, which makes it useful for editing and exchanging presentations across tools. It is still in early development, so complex formatting, embedded media, and advanced PowerPoint features may not always transfer perfectly; important files should be backed up and reviewed before sharing.&lt;/p&gt;

&lt;p&gt;The project keeps files on the user’s device and is released under the MIT or Apache-2.0 license. You can try &lt;a href="https://deckcraft.me/" rel="noopener noreferrer"&gt;DeckCraft&lt;/a&gt; online or explore the &lt;a href="https://github.com/storytold/deckcraft" rel="noopener noreferrer"&gt;source code&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>opensource</category>
      <category>rust</category>
      <category>software</category>
      <category>productivity</category>
    </item>
    <item>
      <title>CADCraft: an open-source Rust 2D CAD workflow for automation</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Sat, 24 Oct 2026 18:55:53 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/cadcraft-an-open-source-rust-2d-cad-workflow-for-automation-41pd</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/cadcraft-an-open-source-rust-2d-cad-workflow-for-automation-41pd</guid>
      <description>&lt;p&gt;When a project needs precise technical drawings and a workflow that can be scripted, &lt;a href="https://cadcraft.org/" rel="noopener noreferrer"&gt;CADCraft&lt;/a&gt; is a free, open-source 2D CAD application built in Rust. It is designed for architects, engineers, students, educators, makers, and technical designers who want a familiar drafting workflow without a subscription or account requirement.&lt;/p&gt;

&lt;h2&gt;
  
  
  A practical drafting core
&lt;/h2&gt;

&lt;p&gt;CADCraft includes command-line tools for common drafting operations such as LINE, PLINE, CIRCLE, ARC, RECTANG, OFFSET, TRIM, FILLET, HATCH, and DIMLINEAR. Users can work with absolute or relative coordinates, distances, and angles. Object snaps, polar tracking, orthographic drawing, autocomplete, command history, aliases, and repeated commands help keep repetitive drafting work precise.&lt;/p&gt;

&lt;p&gt;Drawings can be organized with layers and extended with associative dimensions, multileaders, tables, hatches, paper-space layouts, and plotting features. The application supports DWG and DXF workflows and can export designs to PDF, SVG, and PNG.&lt;/p&gt;

&lt;h2&gt;
  
  
  Built for automation
&lt;/h2&gt;

&lt;p&gt;The interesting part for developer workflows is the built-in automation surface. CADCraft exposes MCP, JSON control, and a headless CLI so users and agents can create drawings, inspect entities, render screenshots, convert files, and verify results. That makes it possible to keep technical drawing steps closer to scripts, repeatable build jobs, and agent-assisted workflows instead of relying only on manual UI actions.&lt;/p&gt;

&lt;p&gt;CADCraft runs on Windows, macOS, Linux, FreeBSD, and the web through WebAssembly. It is released under the MIT or Apache-2.0 licenses and is still in active development, so it is a project to evaluate, inspect, and contribute to as it evolves.&lt;/p&gt;

&lt;p&gt;For anyone exploring open-source CAD with a programmable edge, the &lt;a href="https://cadcraft.org/" rel="noopener noreferrer"&gt;CADCraft project&lt;/a&gt; is worth a look.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>devtools</category>
    </item>
    <item>
      <title>GridCraft: An Open-Source Spreadsheet for Practical Data Work</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Thu, 22 Oct 2026 13:07:42 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/gridcraft-an-open-source-spreadsheet-for-practical-data-work-5gpg</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/gridcraft-an-open-source-spreadsheet-for-practical-data-work-5gpg</guid>
      <description>&lt;p&gt;GridCraft is a free, open-source spreadsheet application created by ArtCraft for practical data work. It combines an Excel-style ribbon, worksheet tabs, a formula bar, formatting tools, charts, and PivotTables in one workspace.&lt;/p&gt;

&lt;p&gt;With GridCraft, users can open, inspect, edit, and save XLSX files while keeping an original copy for safety. The application supports more than 500 worksheet functions, including SUM, IF, dynamic arrays, LET, and LAMBDA. Its calculation engine is written in Rust and runs locally through WebAssembly in the web version.&lt;/p&gt;

&lt;p&gt;GridCraft works in browsers and as a desktop application for Windows, macOS, and Linux. It is distributed at no cost under the MIT and Apache 2.0 licenses. As alpha software, it has compatibility limitations with complex workbooks, specialist Excel functions, external data, and VBA macros, so important files should be backed up and checked after import or export.&lt;/p&gt;

&lt;p&gt;Learn more at &lt;a href="https://gridcraft.me/" rel="noopener noreferrer"&gt;GridCraft&lt;/a&gt; or explore the &lt;a href="https://github.com/storytold/gridcraft" rel="noopener noreferrer"&gt;source code on GitHub&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>opensource</category>
      <category>spreadsheets</category>
      <category>productivity</category>
      <category>rust</category>
    </item>
    <item>
      <title>Ideogram AI 4.5: Sharp Typography, Native 2K Output, and Drift-Free Localized Editing</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Tue, 13 Oct 2026 11:19:32 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/ideogram-ai-45-sharp-typography-native-2k-output-and-drift-free-localized-editing-5a6n</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/ideogram-ai-45-sharp-typography-native-2k-output-and-drift-free-localized-editing-5a6n</guid>
      <description>&lt;p&gt;Ideogram AI 4.5 is an advanced image generation and editing tool built for precise, high-quality visual creation. Its tagline sums up the workflow well: create boldly, edit precisely.&lt;/p&gt;

&lt;h2&gt;
  
  
  What it does
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Sharp typography.&lt;/strong&gt; Text rendered inside generated images stays crisp and legible, which matters for posters, labels, ads, and product shots.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Native 2K output.&lt;/strong&gt; Results come out at 2K resolution without a separate upscaling step.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Localized, drift-free edits.&lt;/strong&gt; Select an area, change it, and the rest of the image stays intact with minimal visual drift - so you can iterate on one detail without re-rolling the whole composition.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Who it is for
&lt;/h2&gt;

&lt;p&gt;Anyone doing iterative visual work: marketers producing ad variants, designers exploring label or packaging concepts, and developers generating assets where consistent, readable text is part of the image.&lt;/p&gt;

&lt;h2&gt;
  
  
  Access
&lt;/h2&gt;

&lt;p&gt;It follows a freemium model and launched on October 8, 2026. You can try it at &lt;a href="https://ideogram-45.com/" rel="noopener noreferrer"&gt;ideogram-45.com&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;If you work with AI image tools, the localized edit that preserves untouched pixels is the part worth testing first - it changes how many iterations a single concept can survive.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>design</category>
      <category>showdev</category>
    </item>
    <item>
      <title>From Two Photos to Fairytale Motion: A Small AI Video Experiment</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Tue, 13 Oct 2026 08:21:59 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/from-two-photos-to-fairytale-motion-a-small-ai-video-experiment-2dfi</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/from-two-photos-to-fairytale-motion-a-small-ai-video-experiment-2dfi</guid>
      <description>&lt;p&gt;Rumpelstiltskin AI is a creative video studio built around a small, simple prompt: start with two photos, choose a setting, add your direction, and generate an original AI-powered dance video.&lt;/p&gt;

&lt;p&gt;Launched on October 8, 2026, Rumpelstiltskin AI is designed for playful, short-form storytelling. The product turns a pair of images into a whimsical fairytale scene, with the user guiding the setting and the creative direction rather than starting from a blank timeline.&lt;/p&gt;

&lt;p&gt;The idea behind the workflow is to make the first step of visual creation feel approachable. Instead of assembling a full production, users can begin with images they already have, add a little context, and see that concept become motion. The result is meant to be a short scene—something you can share, remix, or use as a starting point for a bigger story.&lt;/p&gt;

&lt;p&gt;Rumpelstiltskin AI uses a freemium model and is available at &lt;a href="https://rumpelstiltskinai.org/" rel="noopener noreferrer"&gt;https://rumpelstiltskinai.org/&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;If you are experimenting with playful AI video, what kind of setting would you choose for your own two-photo fairytale?&lt;/p&gt;

</description>
      <category>ai</category>
    </item>
    <item>
      <title>A Practical Guide to Better AI Image Prompts in the Browser</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Sun, 11 Oct 2026 19:08:10 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/a-practical-guide-to-better-ai-image-prompts-in-the-browser-bbh</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/a-practical-guide-to-better-ai-image-prompts-in-the-browser-bbh</guid>
      <description>&lt;p&gt;AI image generation is most useful when the prompt is treated as a small, repeatable workflow rather than a one-shot command. This guide covers a simple loop for creating and refining images in the browser.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. Start with the subject
&lt;/h2&gt;

&lt;p&gt;Lead with the main subject, then add the setting, lighting, and mood. For example:&lt;/p&gt;

&lt;blockquote&gt;
&lt;p&gt;A ceramic mug on a linen tablecloth, soft window light&lt;/p&gt;
&lt;/blockquote&gt;

&lt;p&gt;This gives the model a clear visual hierarchy before you add style details.&lt;/p&gt;

&lt;h2&gt;
  
  
  2. Quote text that must appear
&lt;/h2&gt;

&lt;p&gt;If the image includes a sign, menu, poster, map, or chart, put the exact wording in quotation marks and say where it should appear. That makes the text requirement explicit instead of leaving it as an implied detail.&lt;/p&gt;

&lt;h2&gt;
  
  
  3. Add references with clear roles
&lt;/h2&gt;

&lt;p&gt;For edits or multi-image compositions, name the role of each reference: the jacket from image 2, the person from image 1, and so on. Keeping the references explicit helps preserve the parts you care about.&lt;/p&gt;

&lt;h2&gt;
  
  
  4. Change one thing at a time
&lt;/h2&gt;

&lt;p&gt;After the first generation, make small edits instead of rewriting the whole prompt. A focused request—recolor the wall, remove one object, or change the sky—makes it easier to keep the face, product, or layout stable.&lt;/p&gt;

&lt;h2&gt;
  
  
  A browser-based workflow
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://nanobanana-21.com/" rel="noopener noreferrer"&gt;nanobanana 21&lt;/a&gt; is a fast, simple image studio powered by Nano Banana 2.1. It supports high-quality image creation and editing, readable text, consistent characters, and up to 14 reference images. The workflow is intentionally lightweight: write a prompt, optionally add references, choose an aspect ratio and resolution, then generate and refine without code or a GPU.&lt;/p&gt;

&lt;p&gt;The broader lesson is portable across image tools: describe the subject first, make constraints explicit, and iterate in small steps. That makes results easier to compare and easier to reproduce.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>productivity</category>
      <category>tutorial</category>
    </item>
    <item>
      <title>From Two Photos to a Cinematic Zombie Love Story: Riding the Zombie AI Trend</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Thu, 08 Oct 2026 03:06:37 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/from-two-photos-to-a-cinematic-zombie-love-story-riding-the-zombie-ai-trend-2jhk</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/from-two-photos-to-a-cinematic-zombie-love-story-riding-the-zombie-ai-trend-2jhk</guid>
      <description>&lt;p&gt;The "zombie transformation" trend is all over short-video feeds right now: two photos in, one person slowly turns into a zombie, and the story ends with an unexpected emotional twist instead of a jump scare.&lt;/p&gt;

&lt;h2&gt;
  
  
  What ZombieTrend does
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://zombietrend.org/" rel="noopener noreferrer"&gt;ZombieTrend&lt;/a&gt; is an AI video generator built around this single format. You upload two photos, and it produces a cinematic 9:16 vertical MP4:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;One person gradually transforms into a zombie.&lt;/li&gt;
&lt;li&gt;The other character chooses love over fear and holds them.&lt;/li&gt;
&lt;li&gt;A flashback to their warmest shared memory closes the story.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  Why the format works
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;The transformation gives you a hook in the first second.&lt;/li&gt;
&lt;li&gt;The embrace flips the horror expectation into romance.&lt;/li&gt;
&lt;li&gt;The flashback ending drives rewatches and shares.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Practical details
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Output is 9:16 MP4, ready for TikTok, Reels, and Shorts.&lt;/li&gt;
&lt;li&gt;There's a free preview before you commit to a full render.&lt;/li&gt;
&lt;li&gt;Pricing is freemium, so you can test the format with zero risk.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;If you have been looking for a fresh spin on couple content, the zombie love story format is worth a try — check it out at &lt;a href="https://zombietrend.org/" rel="noopener noreferrer"&gt;zombietrend.org&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>aivideogenerativeaiphotography</category>
    </item>
    <item>
      <title>From a Scene Idea to Short-Form Video with Genjutsu AI</title>
      <dc:creator>WEI LIU</dc:creator>
      <pubDate>Wed, 07 Oct 2026 10:15:08 +0000</pubDate>
      <link>https://dev.to/wei_liu_33dd3614a31eda120/from-a-scene-idea-to-short-form-video-with-genjutsu-ai-3io5</link>
      <guid>https://dev.to/wei_liu_33dd3614a31eda120/from-a-scene-idea-to-short-form-video-with-genjutsu-ai-3io5</guid>
      <description>&lt;p&gt;When a creative idea is clear but the production pipeline is not, a focused generation workflow can help turn that idea into something shareable.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://genjutsu-ai.org/" rel="noopener noreferrer"&gt;Genjutsu AI&lt;/a&gt; is built for that first step: turning ideas into cinematic videos, product ads, social media shorts, and high-quality images in minutes.&lt;/p&gt;

&lt;p&gt;The workflow is simple:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Choose a model.&lt;/li&gt;
&lt;li&gt;Describe the scene you want to create.&lt;/li&gt;
&lt;li&gt;Generate content in a format made for TikTok, Reels, YouTube Shorts, and more.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;That makes it useful for exploring visual concepts, preparing product-ad variations, or moving from a rough scene description to a short-form asset without starting from a blank timeline.&lt;/p&gt;

&lt;p&gt;If you are experimenting with AI-generated video and images, you can learn more at &lt;a href="https://genjutsu-ai.org/" rel="noopener noreferrer"&gt;genjutsu-ai.org&lt;/a&gt;.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>tooling</category>
      <category>design</category>
      <category>startup</category>
    </item>
  </channel>
</rss>
