Creating Beautiful, On-Brand Photography with AI
summary
Shows a working product: Lummi — curated AI stock gallery with search, reframe crops, chutes (consistent series), and retouch features, plus experiments using Midjourney. Demonstrates a concrete prompt method — prompt recipe: subject→action→setting→lighting→film — and gives before/after prompts that change age, ethnicity, tattoos, clothing, and camera look. Notes models bias toward young skinny blue‑eyed defaults, that humans clean up glitches (extra fingers, odd limbs), and that video outputs still fail often.
AI-written summary · about
chapters (9)
- 0:00Introduction and welcome
- 5:23Introducing Lummi AI tool
- 8:31Challenges with AI-generated images
- 10:11Creating inclusive AI photography
- 19:22Crafting effective AI prompts
- 31:12Demo of AI image generation
- 39:28Q&A on AI ethics and sustainability
- 44:37AI risks and alignment problem
- 51:41Closing and karaoke fun
related talks
01h 07mMinimizing Bias in AI-Generated Images
36mArt in the Age of Algorithms
49mCreation as Curation [Keynote]
19mAI-Powered Startup Toolkit for a Product Designer
33mAI Images: Redefining Quality with Innovation
01h 00mPanel Discussion: AI Meets Design: A Dialogue
47mAI Image Generation: Visual Artistry with Midjourney, Adobe Firefly, and DALL-E3
39mVideo Generation + AI: From Production to Product