SEEING IS BELIEVING is an audio visual experiment which explores how information mutates as it spreads. Algorithms and relentless content-feeds distort information into increasingly wild unrealities through a cycle of generation loss. We carry these impressions with us, lacking any knowledge of (or care for) the original source. The imagery presented will not prompt the same reactions from every viewer. The accompanying sounds are unstable and malleable - they contort in response to the images. The visualization and sonification elements of the piece were developed using custom software. Image-to-sound transcoding was utilized as a tool to bridge the gap between disparate forms of digital information. The unpredictable outcomes exemplify the phenomenon of cumulative error in digital spaces. SEEING IS BELIEVING is a custom software-based experiment using algorithmic interpretation of information to generate a distorted dream logic from human content.