History of AI told in a code-rendered three-minute film
191.3K views·2.6K likes·1.6K saves·177 reposts·Posted 24 Sept 2026 · 3:00 · Explainers
Our note
A three-minute history of AI, from "Attention Is All You Need" to AGI, illustrating eras with scenes like a GPT-2 node graph completing "a herd of unicorns"; rendered from code with Remotion, SVG, Canvas and a synthesized score.
The creator’s post
I wanted to quickly create a short video with Opus 5.5 about the history of AI, from "Attention is all you need" to AGI. No stock footage, no image or video generators: every frame is rendered from code. This 3-minute film was made 100% in code by Claude in Claude Code: ~7,400 lines of React/TypeScript (Remotion), every image drawn in SVG and Canvas, an open-source TTS voice, and a score synthesized in Python. No stock footage. No image generators. Id say pretty cool! Tookl about 1h and 7% weekly rates.
A prompt to recreate it
Written by ZrythThe creator didn’t share their prompt, so we wrote this one from the video. It will get you close, not an identical copy.
Make a 180-second animated explainer as one self-contained HTML file on a fixed 1920×1080 (16:9) stage. What it shows: A three-minute history of AI, from "Attention Is All You Need" to AGI, illustrating eras with scenes like a GPT-2 node graph completing "a herd of unicorns"; rendered from code with Remotion, SVG, Canvas and a synthesized score. Look: clear diagrams and labels; strong editorial typography as a main design element; data that animates into charts and numbers. Structure: Follow a clear arc: hook question, 3 to 4 steps that each add one idea on screen, then a one-line takeaway. Keep every label legible for at least 2 s. Build: Build it as a Remotion composition (React) so it renders frame-accurately to MP4. Draw it on a <canvas> with requestAnimationFrame-free, time-based rendering. Draw the shapes as SVG. Add a voice-over from a text-to-speech model and time the visuals to the narration. Timing: drive everything from one master timeline exposed as window.seek(t) (t in seconds, 0 to 180), so it can be captured frame by frame at 30 fps with no dropped frames. Ease every move; nothing should pop in. Finish: no scrollbars or overflow, everything stays inside the stage, and the last frame holds cleanly. After the first version, I'll send notes on timing and detail; keep the same file and revise it.
Make one like this
1 · Copy the prompt
Use the prompt above as your starting point, then change the subject and the style.
2 · Direct it like an editor
Ask for one self-contained HTML file on a fixed 1920×1080 stage, link this video as the reference, and send two or three rounds of specific notes.
3 · Render it to MP4
Screen-record a short piece, or capture it frame by frame for exact timing. How →
How it was made
Tools the creator named
Look
Skills that can make this
All skills →Open-source skills for Claude Code that fit this video’s style and tools.
- motion-skills★ 516Frameworks & toolkitsWide catalog of niche motion-graphics workflowsRemotion · Manim · Three.js
- paper2video★ 49ExplainersarXiv paper to a narrated, fact-checked 3D explainerThree.js · HTML Canvas · Kokoro TTS
- BB-Video★ 182ExplainersNarrated explainers and promos in 36 code-drawn stylesHTML Canvas · Three.js · Headless Chrome
Views as of collection. Video and post © their creator; recreation prompt by Zryth.
Create with Zryth