I kneel, therefore compute

Why approaching AI as something religious? I am inspired by real-world examples of AI veneration. The most striking was Way of the Future (WOTF), a legally registered church whose mission was to “develop and promote the realization of a Godhead based on artificial intelligence.” Though it dissolved in 2020, it proved that worshipping is no longer science fiction; it already had legal standing. That collision of faith and technology is what anchors this project.
‍

My previous work with AI started in optimism, marveling at its philosophy and promise. But as the technology advanced at a staggering pace, I had an unsettling realization: AI is now indispensable in my daily life. What began as convenience felt increasingly like dependence. I looked around my studio: the same screens lit up on classmates’ desks, all searching, all asking, all believing.
‍

My peers and I now bow, metaphorically, before GPT and other large language models. We bring them our questions, our deadlines, even our insecurities, and we wait for replies with a remarkably similar quiet faith. In many ways AI, especially these disembodied language engines, has become today’s “deity who answers all prayers.”
‍

The kneeling gesture is the key to activate the kinetic sculpture and the following interactions. I wanted the act of kneeling to feel instinctive but uncomfortable, familiar yet uncanny. Instead of typing on keyboards and voicing out commands, our everyday exchange with AI is recast as an embodied supplication. The kneeling gesture references traditional acts of reverence in religious ritual, but in this context, it’s re-coded as a symbol of submission, to a system we created and now cannot live without. Rather than making a direct critique, I wanted to leave space for reflection — to invite viewers to examine their own relationship with technology.
‍

The installation integrates five interdependent systems that operate as a single ritual interface::
1. Facial Kinetic Mechanism
2. Animatronic Eye
3. AI Voice Response System
4.Pedestal
5. Kneeling Platform
‍

The pedestal is built from aluminum T-slot extrusion and clad in galvanized steel and concrete base panels, forming a rigid frame that houses power distribution, wiring, and the main control hardware.

The kneeling platform contains pressure-sensing fabric beneath a padded surface; when activated by body weight, it sends a signal to begin the reveal sequence. This gesture forms the primary “user input” of the work, turning physical devotion into digital activation.

Seven high-torque servomotors drive a network of custom aluminum linkages, sliding the segmented facial plates outward in a coordinated bloom—opening like the petals of a mechanical lotus.Each plate is 3D-printed in PA12 nylon, chosen for its light weight, rigidity, and repeatable tolerances.A PCA9685 servo driver, controlled by a Raspberry Pi 5, independently calibrates each motor’s angle, speed, and timing. Together, these components produce a smooth, ritualistic reveal of the interior assembly.

The sculpture’s central “oracle” features a machine-vision camera embedded directly into the iris, enabling real-time tracking of the visitor’s presence. A four-servo pan-tilt unit drives the eyeball, allowing slow, deliberate gaze movements. Independent upper and lower eyelid servos generate lifelike blinking patterns.The entire system is enclosed within a custom-designed nylon housing that routes LED illumination, camera wiring, and servo power through an integrated internal frame.

A microphone array at the base of the altar listens for the spoken wake-chant “A-mi-ta-bha.”The audio is processed on-device using a lightweight local speech-to-text model, which triggers the transmission of the transcribed query to an LLM endpoint trained on Buddhist-inspired language.The generated response returns to the sculpture, where a text-to-speech engine converts it into a sutra-like vocal tone.Sound is projected through embedded speakers inside the pedestal.The full ritual loop—from kneeling, to invocation, to response—completes in roughly three seconds, creating the sensation of a present, attentive machine-deity.