Dark Matter, 2021
Dark Matter is an installation (50in*26in*46in), which is a matte black upward facing Closed-circuit television (CCTV) camera shaped wooden sculpture. It has a reactive dark liquid “lens”, displaying pixelized low resolution surveillance footages. A computer monitor is placed facing to the sculpture, playing a 3m41s multi-channel video in loop. Three of these channels have a common source of edited found footages but with different artifacts, and the most distorted one controls the movement of the liquid lens surface, creating a mirroring effect. The fourth channel plays a morphing video of CCTVs, it is created by StyleGAN machine learning algorithm based on a custom dataset.
The material in the CCTV’s lens, is called Ferrofluid, which is a liquidated reflective magnetic fuel that reacts to magnets. It was invented by NASA engineer, Steve Papell in 1960s, in order to control fuel’s movement in zero gravity. The physicality of ferrofluid is similar to dark matter, they both have invisible forces which interact with visible substance.
The increasing industrial use of artificial intelligence training in large datasets became a security and privacy issue. Artificial intelligence technologies, such as body tracking, facial recognition, user behavior analysis, object detection, etc. are mainly used by authorities and companies as a predictive tool for global surveillance and customized targeting ads. People are leaving foot traces everywhere and giving out free data constantly. Tech giants and telecom companies built massive amount of submarine cables across the globe, which creating a network of communication and information exchange. According to Telegeography, as of early 2021, there are approximately 426 submarine cables in service around the world. Based on their map, about one-third of these internet cables are owned by Google, Amazon, Facebook and Microsoft. Surprisingly, approximately half of these tech companies owned undersea internet cables are connecting the west coast of the U.S. and the east coast of China, even though the publics in mainland Chinese do not have direct access to Google, Amazon and Facebook.
Based on a HIS Markit report in 2016, China is expected to have 567 million CCTV cameras in use in 2021. Chinese government created a monitoring system called Skynet, which uses AI technologies to monitor prevent crimes. More advanced models can detect and recognize people when they are wearing a mask due to the Covid-19 pandemic. Furthermore, algorithm of identifying people from their walking style is being developing. All of these machine learning algorithms need massive amount of data to train. Therefore, a village in China called Guizhou became a central factory for pre-process data. Farmers, students, etc. are now working for machine to label objects in images. AI seems to be very smart and take over people’s job, however, it utilizes more energy and laboring work.
The development of technology is not invertible. The machine or the algorithm itself has no intend to benefit or harm, rather those
who controls the tool decides how to pressure the use.