Send message to Contributor

Short abstract:

A speculative fictional parody device that claims extra-sensory perception (ESP) capabilities, reading thoughts from cognitive radiation. The MindReader ESP32 reflects on surveillance systems that are based on capturing background radiation emitted from devices, such as Bluetooth space occupancy systems, which can surveil without notice or consent.

Abstract:

The MindReader ESP32® by NeuralChain AI™ is a speculative fictional parody of an AI-enabled device that claims to have extra-sensory perception (ESP) capabilities. It purports to read minds based on inferring patterns in electromagnetic radiation emitted from the brain during specific cognitive acts. This physical prototype is built on an ESP32 microcontroller, a widely-used low-cost wireless microcontroller. It is equipped with a small screen that scrolls randomly generated descriptions of fictional EASST/4S attendees' presumed thoughts and emotions. The device queries Mistral's Mixtral, an open-source large language model (LLM) hosted on a private server, to generate humorous and self-referential descriptions of attendees' mindsets. The MindReader ESP32 serves as a commentary on the increasing use of surveillance systems that are based on capturing background radiation emitted from devices, such as Bluetooth beacon space occupancy systems, which can surveil without notice or consent. The MindReader ESP32 displays scrolling ‘live’ fictional data based on attendees' emotions, interests, and thoughts, such as: {Excitement: high; Curiosity: moderate; Nostalgia: low; Thoughtfulness: high; Contemplation on the blend of technology and society; Awareness of diverse perspectives in the room; Anxiety about interactive poster session; Anticipation for engaging conversations; Hunger: medium;"} While the data is entirely fictional, it invites participants to consider the implications of real-world systems that collect and analyze personal information. This hands-on demonstration aims to spark discussions on the implications of such ways of knowing on privacy, consent, transparency, regulation, and design principles. This references earlier work such as the CHI 2014 Quantified Toilets performance art.