is an autoethnographic exercise that took place between 2015 - 2022. It began by collecting intimate sound fragments with the intuition they would be yearned for. This is an attempt to visualize them in a state of pre-meaning, where the voice memos are sonically abstracted.

This kind of retrospective was made possible by W-GAN neural networks and custom interpretation workflows.

Special thanks to Jorge Lorenzo Flores and all the memory co-creators.

⇿

Click play and navigate between real memories (left - Real) and reconstructed ones (right - Fake).