43 design principles to kill the generic AI look in generated images
Kaihzu · reddit · 2026-10-28
A designer-turned-AI-experimenter shares 43 principles for fixing the instantly recognizable "AI look" in generated images, developed by treating ChatGPT like a junior designer and critiquing specific problems instead of re-rolling.
Key takeaways:
- Decluttering ≠ changing style: criticizing a neon-heavy poster just gets you an equally cluttered vintage poster — fix the underlying problem, not the aesthetic
- Restraint over spectacle: one dominant focal point, deliberate negative space, remove non-contributing elements
- Typography: legibility first, limit competing treatments, build a clear reading order
- Don't invent details: fake addresses, dates, QR codes and unrequested promotional copy should go
- Avoid AI clichés: neon, glowing particles, lens flares, cinematic everything; polish isn't originality
- Treat critique as art direction: preserve what works, don't auto-agree with every criticism
The guiding philosophy: design should look intentional, not merely impressive. Before/after poster comparisons included.
More from Multimodal
- Seedance 2.5 hailed as taking AI filmmaking to another level — SimplyAnnisa · 2026-10-28
- Hands-on with Baidu's Unlimited OCR 2: whole docs in one pass, no page limit — Baidu_Inc · 2026-10-28
- Agentic video editing goes pro: senior editors 5x output, costs to collapse — dotey · 2026-10-28
- Alibaba's OvisOCR3: a 0.8B model tops OmniDocBench at 96.74 via multi-teacher distillation + RL — ATH-MaaS · 2026-10-28
- CodeCap-Bench: executable code beats natural language as video captions — Jiafu Tang · 2026-10-28
- Moving Atoms Engine turns real-world spaces and cities into high-resolution 4D models for sim — Scobleizer · 2026-10-28