The Looking Glass: AI Engine Generates Historical Images of Any Global Location
josh_bickett · x · 2026-08-01
Plinius introduced 'The Looking Glass,' a spatiotemporal AI image and video engine. Users can select any point on the globe at any time (past, present, or future), and the tool uses multimodal models to generate an image of what it imagines was happening there at that exact hour, with an option to bring it to life as a video. The creator notes it's not perfect science yet, but quality will improve as models and prompts progress.
More from Multimodal
- Higgsfield Teases Upcoming Video Model Seedance 2.5 — tisch_eins · 2026-08-01
- Reve 2.1 vs Krea 2 Turbo: A 192-Image Side-by-Side Comparison — dh7net · 2026-08-01
- ComfyUI Tip: Fast Qwen VL Prompting Without Extra Plugins — xbobos · 2026-08-01
- AI Video Demo: First-Person Sprint Through Endless Surreal Rooms — fofrAI · 2026-08-01
- Chimera: Bringing LLM-Style Architecture and Scaling to Visual Generation — teortaxesTex · 2026-08-01
- Claude + Thrixel Generated a Full 3D Game with Under 5 Prompts — RanaHanocka · 2026-08-01