[help me know the truth]


[help me know the truth]

Custom software, 4 tablets, network, projector, dimensions variable

[help me know the truth] is an installation that uses computational neuroscience algorithms to reveal people’s collective unconscious biases. I use the “reverse correlation” method in computational neuroscience to change images and ask visitors to choose one of the altered images based on a prompt.

Visitors at the gallery first take a selfie. The software adds noise algorithms to the face, producing two versions of the selfie that are slightly different. Then visitors are asked to ‘judge’ the face with a randomly assigned prompt: to choose which face is more trustworthy, for example, or less friendly; which is nicer, or likely a terrorist, etc. The noise pattern is saved with the selected face and the process progresses, moving from iPad to iPad around the space. The resulting image after being altered and judged multiple times finally appears on a large projection, the face ‘manifesting’ our otherwise invisible collective biases based on the prompt.

That gallery visitors are constantly judging the people around them (the selfies are distributed immediately around the gallery) and that other people in the gallery are judging them is a key part of the enjoyable, yet also uncomfortable, experience of the work.

The intent behind [help me know the truth] is to both utilize and question how computational neuroscience techniques can uncover the categorizing systems of the mind, and how they are therefore subject to imperceptible, socially constructed fears and values. The work further calls into question notions of truth. Is neuroscience truly the best way to parse the complex interplay of thoughts, feelings, and beliefs of a person? If not, do we trust computers to categorize our physical appearances? Do we trust ourselves?



Ars Electronica Press Release. “Future Humanity – Our Shared Planet” September 2018.

Ars Electronica Catalog “CyberArts 2018

D’Auria, Veronica. “[help me know the truth] Mary Flanagan (US)-a software-driven participatory artwork for Ars electronica.” 7 October 2018.

Additional Credits

Jared Segal, Ron Dotsch whose reverse correlation software was adapted for this project, Danielle Taylor, and Sukie Punjasthitkul