The Turing Test

Inspiration

Real bars vs recycled noise. Heart vs hardware. Welcome to the Turing Test. After a discord rap battle, the artist Ghost Ryder and me, we knew we wanted to collaborate, but only when the concept, setup, and execution were right. The idea was: We wanted to draw a clear line between artists who guide AI with intent and taste, and people who push out algorithm slop for clicks. This project is also a calling card for AIU.FM (online AI web radio), our curated home for AI-assisted music made by humans who care.

What it does

It stages a musical Turing Test. The video opens with a generic, corporate-sounding AI rapper to set the baseline. Then we answer with crafted verses, beat switches, and visual shifts that make the difference obvious. Viewers hear and see how direction, story, cadence, and editing separate human-led artistry from zero-effort generation. It entertains, makes a point, and invites people to check out our home at AIU.FM.

How we built it

Concept and writing: We outlined a three-act structure that moves from generic to intentional. Lyrics by Aidan Yagu, collaboration with Ghost Ryder, structured for tight flows, internal rhymes, and a clear narrative arc.

Music production: Backing tracks generated in Suno with focused prompts and multiple iterations. We mapped sections to bar counts, refined tempo and transitions, and arranged ad-libs and call-backs to emphasize the thesis.

Visual pipeline:

Storyboards and style frames in Leonardo.ai to lock look, lenses, and camera paths. Video sequences in Sora for key scenes. We used shot-specific JSON prompts, kept character continuity, and controlled motion language to avoid physics drift. Post in DaVinci Resolve: color, timing to transients, speed ramps, match cuts, selective motion blur, subtle glitch overlays, and brand consistency for the Aidan mask with the vertical line. Project management: Shot list and versioning, prompt library, and a shared review cadence to keep audio and visuals locked.

Challenges we ran into

Sora’s tendency to default to realism or warp physics on fast movement, which we solved by tightening prompts to keep camera directions steady and adding post motion cues or let it hallucinate on purpose to see what we can use. Character consistency, especially the vertical face line, which required prompt refinements and lots of frame curation by using first frame last frame setups Timing drift between AI-generated instrumentals and vocal delivery, solved with bar-accurate editing and micro time-stretching.

Accomplishments that we're proud of

A concept that actually demonstrates the difference between tool use and tool worship. Verses, visuals, and edits that hit on the beat and serve the message. A cohesive aesthetic that matches my Aidan Yagu universe and the AIU.FM identity.

What we learned

Direction beats randomness, but not always. Clear storyboards and bar maps save time and raise quality, but Sora sometimes surprised us with with random scene shifts I actually kinda liked Still: AI tools are amplifiers. If you bring craft, they help. If you bring nothing, they show it.

Small technical tweaks matter: motion hints, depth cues, and restrained post effects sell the shot better than brute force prompts. Collaboration works best when feedback is specific and tied to the narrative goal, not just the look.

What's next for The Turing Test

honestly, I haven't though about that yet.... I mainly made this together with Ghost as a sort of welcome video to our discord community and AI radio. It's sitting now on my Youtube account waiting for a random algorhythmic breakthrough :D

Credits: Lyrics/Production: Aidan Yagu Featured Lyrics: Ghost Ryder Music: Suno Storyboard: Leonardo.ai Video material: Sora AI Editing: DaVinci Resolve

Built With

• davinci
• leonardo
• sora1
• suno