The Video Surveillance Debate

It seems an interesting conversation is going to have to start taking place in many democracies. Machine learning has reached a place where it is ready to be deployed for facial recognition, threat assessment, and machine intelligence based surveillance. Some of our readers who live, or travel to China frequently, know this technology is already mass deployed in nearly all major cities where video cameras are prevalent. For those of you who are not terribly familiar with some of the implementations of how China is using this, a quick story.

A friend of mine travels to China often to recruit students for the academy where he is the head of school. On a recent trip, he and his local associates went to buy train tickets the day before he had to travel inland. When he returned the next day to board the train, he was denied entrance because his ticket was for the day prior (the day he purchased them) not for the day he intended to travel. Trying to sort out the mistake, he went to the ticketing camera and explained the ticket he purchased was supposed to be for today and not yesterday as the operator claimed. The sales associate behind the counter asked him to wait a moment and went into a back room. He came out moments later and showed my friend video of him and his translator purchasing the tickets, with caption translation, and confirmed the error in the translator who did, in fact, say to buy the tickets for the prior day travel, not the day he intended. He was telling me this story and was astonished at the speed in which the sales associate found the video of him purchasing the tickets and brought back just that snippet of video to confirm the error was not the sales associates fault.

There is no doubt the sales associate was able to use the current video of my friend to search for the video the day before to get the recorded account of his transaction in order to sort out what actually happened. As you can imagine, this sort of thing will make a lot of Westerners very uneasy. But we should have a more open conversation about the benefits of having more intelligent surveillance and what sort of regulation needs to be implemented in order for the West to use this technology safely and to the benefit of its citizens.

Progress is Always a Trade-Off
Transitions from the old to the new are never terribly easy at a societal level. There is always a demographic who prefers the old way to the new ways, and either does not see the trade-offs or, most likely, doesn’t feel the trade-offs are worth it. This is all ok and normal, and we have a great deal of human history to rely on to understand these lessons.

What makes the video surveillance and machine intelligence layer interesting is at a fundamental level, when we go into public spaces, we are already being surveilled and have been for some time. Nearly all major public spaces have had surveillance cameras for years as a matter of store safety, insurance in case of robbery or theft, theft prevention, and a range of other reasons. What’s new in this equation are systems that can identify individual people and track them accordingly. This is likely where most of the debate and regulation is likely to focus.

If you were to ask any normal citizen, they would probably be OK with surveillance intelligence that was able to do threat detection, threat prevention, and a host of other things that apply to public safety. Where people will likely get more uneasy is when they can be individually identified. Here again, there are trade-offs and benefits, and this is likely where some regulation will have to set in.

At CodeCon Amazon’s AWS CEO Andy Jassy made mention of Amazon’s sale of their facial recognition technology solutions (Recognition) to the US Government and defended their position. However, he also went on to explain why regulation is needed, and he is in support of said regulation around facial recognition technology for video surveillance. Jassy’s position is one of understanding why there are concerns but does not feel there are not benefits to such technology, and he does not feel we should immediately ban or condemn new technology.

I tend to agree with Andy in that there are great benefits, especially around public safety and accountability, that will be found valuable to countries citizens. That being said, one could argue that while government regulation is needed or necessary, that any technology in this category should be run by a private company instead of any government or state. I say this primarily on the point I made about accountability and how this technology can and would hold not just public citizens accountable but also government and state agencies. My hunch is that a private organization that is protected from government or state influence is the better organization to manage and ensure the protection of citizens when it comes to this kind of technology.

Another idea is to develop a consortium of private companies that could include companies like Amazon, Apple, and Microsoft, (maybe Google but I know how people feel about them) and this consortium would be responsible for responsibly deploying this technology and protect the interest of the public.

I bring this up because it is now we have to start having these conversations and start developing a plan before it gets out of hand or too hard to reign back in control from the wrong hands.

Published by

Ben Bajarin

Ben Bajarin is a Principal Analyst and the head of primary research at Creative Strategies, Inc - An industry analysis, market intelligence and research firm located in Silicon Valley. His primary focus is consumer technology and market trend research and he is responsible for studying over 30 countries. Full Bio

Leave a Reply

Your email address will not be published. Required fields are marked *