The government of China is getting ready to set up around 20,000 cameras throughout the city of Shenzen, the NYTimes reports, and the city already has 180,000 owned by the government and private companies. They’re not as much cameras as they are eyes, though, as they’re part of a plan to use AI software to constantly track and identify people’s faces and unusual activities in hunt for police suspects and crime. The police will have the right to link pre-existing cameras at any time into the system. Two-hundred-thousand cameras! My AI science enthusiasm screams “Interesting!” but my human alter-ego simply screams.

Additionally, they’re issuing ID datacards to most citizen that contains work history, educational background, religion, ethnicity, police record, medical insurance status, landlord’s phone number and personal reproductive history. Apparently “I see you’ve had 3 kids, you Kebab-making, South-Korean, sick and in-debt, buddhist, undereducated criminal! Have you paid rent this month? I’m calling your landlord!” will be something Chinese police officers might be blurting out next time they pull people over for a routine check.

Shenzhen, a computer manufacturing center next to Hong Kong, is the first Chinese city to introduce the new residency cards. It is also taking the lead in China in the large-scale use of law enforcement surveillance cameras — a tactic that would have drawn international criticism in the years after the Tiananmen Square killings in 1989.

Well, I’ll say. I’m not the one to oppose AI driven surveillance cameras, but in a country like China I can’t help but wonder how much testing this system has gone through. Not to mention what standards they’re adhering to when it comes to its usage. Will any police officer be able to access the AI systems? How will security be maintained and what precautions are being taken to ensure no unauthorized persons gain access? And that’s actually what the Human Rights Watch in New York is concerned about; China has fewer limits on police power than Western Civilizations.

With a system that’s awake and capable of tracking people through the 200,000 eyes — I wouldn’t want a human capable of meddling with the system in his free time to track anyone he wants for the fun of it.

New York police announced last month that they would install more than 100 security cameras to monitor license plates in Lower Manhattan by the end of the year. Police officials also said they hoped to obtain financing to establish links to 3,000 public and private cameras in the area by the end of next year; no decision has been made on whether face recognition technology has become reliable enough to use without the risk of false arrests.

What? I’ll tell you; there’s always a risk of false arrests. Unless you’ve made a system smarter than us humans (which’d be awesome), as even humans constantly make mistakes. But that’s a bit harsh of me, the writer obviously meant “reliable enough to use without too much risk of false arrests”.

Personally, I don’t see much wrong with using AI software to track faces, provided that they’re used through highest security standards to prevent humans from abusing it. There isn’t any difference in having the eyes of police officers on the street than the eyes of one’s and zero’s. If anything, it’s better, because what the system sees doesn’t need to be revealed unless criminal activity is detected and hence our privacy gets better protection. A human police officer can’t surpress his memories. But again, privacy concerns depend entirely on security and limits of human access to the system. We’re at the beginning of an AI era. We need to be careful to not have our new toys abused.

Links & References

• China Enacting a High-Tech Plan to Track People, NYTimes