zero-degree machine

[2026] Scan to Destroy

Scan to Destroy explores the fragility and volatility of digital infrastructures and their perpetual negotiation with entropy. The QR code is one such infrastructure.
Commissioned by Diatribe Records to explore the limits of ubiquitous infrastructure, this work asks a simple but unsettling question: how much destruction can a system tolerate before it collapses?

Scan to Destroy explores the QR (Quick Response) code as a fragile yet foundational infrastructure. Its usability lies in its resilience, using Reed–Solomon error correction, a technology developed by NASA for deep space exploration in the 1970s. This algorithm keeps codes scannable even when damaged. In recent years, the QR code has shifted from a disposable shortcut to a key piece of global economic infrastructure. In China, it plays a central role in everyday life, serving as a near-universal interface for payments and authentication.

In 2014, WeChat and Alipay popularised digital ‘red envelopes’ during Chinese New Year, accelerating the adoption of QR-based payments across daily life to the point that paying for goods and services by other means became increasingly difficult. Users could send money to a group chat. The app would “randomly” distribute the total amount among friends. Giving money was gamified. Users became obsessed with “snatching” the biggest portion. To claim the money, users had to link their bank cards to WeChat. In one weekend, WeChat did what would have taken years of marketing: it convinced millions of people to trust a social media app with their banking information.

Once millions of people had money sitting in their digital wallets, they needed a way to spend it. The solution: QR codes. WeChat and Alipay sent teams of sales agents out with bags of paper QR codes, teaching shop owners how to use them. Infrastructure made of paper now underpinned one of the world’s most powerful economies. The fragility of this new economic infrastructure required new technological innovation. QR-scanning software such as WeChat’s built-in scanner was designed to correct errors in damaged codes to a degree that exceeds many common Western scanners, including those on the iPhone.
At the centre of the installation is an algorithm that pixel-by-pixel destroys QR codes. The algorithm preserves scannability for as long as possible by using WeChat’s neural network inference. This process unfolds in parallel across three screens.

A QR code undergoes destruction. The process is visualised by a spectrogram, and a live terminal log records the algorithm’s decisions.
The destruction is a negotiation between the algorithm and the image, between order and entropy, and between legibility and noise. Sound fills the space as destruction is rendered audible, turning visual decay into a temporal experience.
At the end of every destruction process, a thermal receipt printer produces the results. The prints accumulate, forming a growing archive of near-failures. The codes remain at the edge of usability. Visitors may attempt to scan them, directly participating in the test of the system’s limits.