I Built an AI Lighting Desk That Understands What You Say

Can an AI understand a creative lighting instruction and turn it into real DMX control? In this video, I demonstrate a prototype AI lighting desk built in Unreal Engine. Using simple text and voice prompts, the system can interpret instructions such as “make the spotlights blue”, “spot the cone” or “create a jungle scene”, then calculate how the available fixtures should respond. This is not simply an AI generating a picture of a lighting design. The prototype outputs real DMX data and understands the fixtures, their addresses, their positions within the virtual scene and how they are already being used. The demonstration includes: • Controlling lighting through natural-language prompts • Sending commands through a cloud-based AI API • Generating real DMX output • Calculating pan and tilt to illuminate objects in the scene • Creating colours, movements and effects • Transferring the Unreal Engine fixture patch into a lighting console • Saving looks as palettes, effects and cues • Separating colour and movement information for reuse • Designing a complete lighting state from a voice prompt The AI-generated results can be transferred into a more conventional lighting-console workflow, where a programmer can edit, refine and play them back. The intention is not to replace lighting designers or programmers, but to provide another creative and technical tool, much like effects engines, macros and timecode systems have done before. This could make sophisticated lighting control more accessible to small venues, educational settings and users who may not have the time or experience required to learn a traditional professional lighting console. This is an early proof of concept, and I’ll continue developing it to explore what AI-assisted lighting design and programming could become. #LightingDesign #StageLighting #ArtificialIntelligence #UnrealEngine #DMX #LightingConsole #CreativeTechnology #LiveEvents #ShowTechnology