Hacking AI for creative operations at scale
summary
Shows concrete methods: uses Figma Weave node flows to turn inputs into repeatable outputs, builds a prompt deconstruction formula, encodes rules as system prompts, and strings nodes into workflow building pipelines. Demonstrations include generating a font in ~30 minutes, a compositor that produces studio-ready product shots for BarkBox, and scale/localization at SharkNinja. Workflows produce batches (router + iterator nodes), claim 95% stability when structured, and swap models without breaking the system.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:58Systematic AI mission
- 3:44Creative inflation challenges
- 5:07Core AI skills for ops
- 8:37Deconstructing media for prompts
- 11:47Building system prompts
- 13:11Workflow in Figma Weave
- 15:20BarkBox composites workflow
- 17:033D mock-ups to studio shots
- 19:10Scaling Shark Ninja localization
- 20:34Key takeaways
related talks
27mPixels and prompts: building products in the AI era
25mDesigning at the speed of AI
33mThe Dawn Of Universal Creativity
19mBeyond agents: AI as a creative partner
18mAI for UI
25mCrafting your AI play(book)
45mHow to Really Scale AI: Combining Traditional and Agentic Automation
17mDesign systems in an AI first ecosystem