2026 / Web Application
Lynch Dream Machine
A web application where users enter a dream, receive a Lynchian image of it, and get a symbolic analysis grounded in Jungian psychology.
- Flask
- Python
- OpenAI API
Overview
Project description
This web application analyzes dreams through a model guided by Jung's analytical psychology, treating dreams symbolically rather than as straightforward narratives. Its interpretive logic focuses on archetypal figures, emotional tone, unconscious tensions, and themes of individuation.
By combining Jungian psychology with David Lynch's world, the system is constrained toward symbolism, ambiguity, and speculative meaning. The result is both a written interpretation of the dream and an image rendered in a Lynch-inspired visual language.
Process
Process
I built this web application in Flask and shaped its behavior through prompt design, using Jungian dream analysis as the interpretive framework. The process involved testing how text generation and image generation could work together so that users would receive both a symbolic written reading and a Lynch-inspired visual response. There was an issue that every dream imputed would flag the image generation model safety guidelines, so a middleman interpreter was added to extract key terms and ambience of the imputed dream to significantly reduce this. The website design was also developed through an iterative process for aesthetics, user navigation, and interactability.
Interface image
Still image from the live project interface.
