What is it?
This artwork responds to faces and spoken words.
Its camera recognizes faces and splays them into rotating flower-like shapes; a central microphone listens for speech and shows its transcripts on a multitude of small screens.
In this work, I am analogizing the data that we spread when interacting in life to pollen: millions of invisible particles flowing in all directions with uncontrollable effects.
“Pollination” uses “whisper.cpp,” a neural network (“AI”) speech transcription tool, to transcribe audio entirely within the device; nothing is uploaded. Its image recognition is powered by OpenCV.
Neither technology is perfect; the same is true of other real-world applications. What happens if a real-time transcript of audio captured in a public place mishears your discussion of a broken icemaker as the name of a terrorist group? Or if a CCTV’s live facial recognition security product misidentifies you?