How Lili Di’s “Like Me” was made: an AI music video without camera, s…
Behind the release of “Like Me”: one creator, a full AI workflow, and a polished pop video produced entirely inside Lilidi.
By Lili Di
There are two ways to talk about “Like Me” by Lili Di . The first is emotional: it''s a pop song about the ache of waiting for a like instead of a real message. The second is practical: it is a fully formed music video built without a camera crew, set, or studio day. [[youtube:https://www.youtube.com/watch?v=o9qIexJcLC8]] That second point is what makes the release especially interesting for creators. “Like Me” shows what happens when image generation, video generation, music creation, and artist world building are handled inside one AI workflow instead of being stitched together across a dozen disconnected tools. One concept, one workflow The release was created with Lilidi.ai , which combines flagship models for image, video, and music in a single studio environment. Instead of treating each asset as a separate experiment, the project holds onto one consistent identity — the same
performer look, the same color world, the same emotional tone. That consistency is the real breakthrough. Many AI first music projects still feel like bundles of interesting outputs. This one feels like a release package. What the process proves A solo creator can now move from idea to finished video much faster than a traditional production schedule would allow. There is no location booking, no glam call sheet, no wardrobe logistics, no reshoots because the weather changed. The time shifts away from physical production and toward direction: prompts, references, iteration, selection, sequencing, and editing. That doesn''t make the work “automatic.” It just changes the labor. Taste matters more. Editorial judgment matters more. The creator is effectively acting as songwriter, producer, art director, video commissioner, and post team all at once. Why music videos are a natural fit for AI
production Music videos live on mood, iconography, repetition, and stylized exaggeration. That makes them especially compatible with generative systems. When the idea is clear, the tooling can produce a surprising amount of polish in very little time. In “Like Me,” the neon city background, glowing heart motif, glossy red styling, and direct performance framing create a world that is easy to recognize across multiple frames and promotional assets. It is exactly the kind of concept that benefits from AI''s speed. From experiment to release strategy The project also works as a release blueprint. There is a full video, multiple thumbnail ready artworks, bilingual positioning, and a clean product story around how it was made. That means the same production can power YouTube, blog content, social clips, launch posts, and press angles. For independent artists, that matters almost as much as
the song itself. The bottleneck is often not making one good asset. It is making enough strong assets around the release to create momentum. Watch the result “Like Me” is a useful example of how quickly the gap between idea and finished music video is collapsing. Watch the full video here: YouTube: https://www.youtube.com/watch?v=o9qIexJcLC8 Built with Lilidi.ai: https://lilidi.ai