FitnessJesus V3
Three data streams – one prophet: church images, song lyrics, and video scripts merge into an evangelical gym podcast.
The Logic
01. Church Image Pipeline: Images from the UdK cloud are loaded and analyzed via Gemma3 Vision. Adjectives describe mood and theme – producing a guiding motif.
02. Song Lyrics Pipeline: In parallel, song lyrics are loaded from the cloud and analyzed for mood/tonality: melancholic, aggressive, hopeful? The emotional intensity is measured.
03. YouTube Pipeline: YouTube titles from the channel "Tomatolix" are extracted. "AlphaJesus" – a fictional alpha-male gym influencer with a man bun and 200k followers – reacts to the titles and hallucinates what happens in the video.
04. Podcast Script: From the church image motif and persona profile, the "Podcastautor" writes a complete FitnessJesus script. Structure: "By our Holy Father, screw them!" → Backstory → Break → Reaction → Community → Outro.
05. TTS + Upload: The script is synthesized to audio via the UdK TTS server and uploaded to NextCloud.
[ JSON Workflow: FitnessJesus_V3.json | 21 Nodes | 3 Pipelines: Vision + Songtext + YouTube ]
Get Blessed
Enter a YouTube topic or keyword:
The "Beautiful Nonsense"
FitnessJesus shows how easily AI can mimic the repetitive, high-energy language of influencer culture. Three completely independent data sources – a church image, a song, and a YouTube video – are fused into a single choleric monologue. "Blessings upon your gains."