Or is it a surveillance system? That’s what most of the press has been calling it since the patent was discovered recently. Is the future of commerce sinister?
Most of the articles are a bit sensationalist, but Buzzfeed does a good job of exploring some of the deeper details here. Just because a patent has been filed doesn’t mean they’ll necessarily deploy it, of course.
- Walmart patented an audio “surveillance” technology that would allow them to generate NLP and sound analysis analytics at the checkout to improve customer experience and manage lineups.
- The very practical business idea (just an idea) created a bit of a stir in the media due to the privacy implications of out-of-home audio recordings of conversations, and the notion that anything an employee says might be, somehow, used against them in the future.
- This is a bit creepy, but we’re also installing AI-driven audio and video surveillance systems in our homes in the form of smart speakers, NLP-powered phones, laptop cameras, PS4 VR depth sensing cameras with voice and facial recognition, etc. – so what gives?
- The key takeaway is that culture is radically changing vis a vis data privacy on a generational level – unless there is a catalyzing event, North American consumer culture will continue to care less about information privacy because the change is gradual and the benefits outweigh the concerns.
- This giving-up of privacy creates more data availability, which in-turn accelerates the power of machine learning systems, as they use data for fuel.
The patent raises a number of interesting topics related to business, technology, and culture as we enter an AI-powered era.
You can imagine the thought process of the creators, which is brilliant. A brand is already able to continuously quantify the quality of their phone, chat, and email-based interactions with customers through AI, at least to a degree, by applying machine learning to the extracted text for emotion analysis, keyword identification (“What % of our CSRs are mentioning ‘ShinyGizmo Upsell’ this month?”), etc. Obviously, there are a lot of elements that aren’t AI-powered in this process, but it’s a critical element of the modern systems doing this work today.
The point of that activity is to ensure that customers are getting consistently high-quality interaction, and that elements of business operations are optimized, like call duration. (My favorite philosophy on average call times is from Zappos, which encourages employees to keep chatting with customers for as long as they like, helping the customer with anything they need – even if it has nothing to do with Zappos.)
So, there’s a gap in information coverage at the dashboard view for the executive team around retail customer service quality. When you’re steering a ship that is as large as China’s entire army, that kind of awareness is critical.
This idea fills that gap. It’s one of those elegant inventions that, if you look at the world through the lens of a marketer, is very cool.
Instead of measuring call-waiting times, it is able to determine based on sound analysis how long a checkout line is, and how fast it’s moving.
Instead of creating “metadata” around a phone session, it’s created around the brand-customer interactions in the physical retail space.
And instead of a manager pulling up a recording of a phone call when a customer makes a dispute, a manager can (theoretically, this is a patent, not a deployed solution) pull up a recording of the conversation.
Perfect! But that’s where things start getting complicated, because instead of a discrete phone call that starts and ends, the system would have to be on all the time. Meaning that instead of “being on” for a few minutes at a time during calls, the employee may feel the pressure to be “always on” in everything they say.
This type of human-behavior analytics has been receiving some very negative backlash at Amazon, where controversy has been brewing around the level of pressure placed on warehouse workers and China has recently rolled out their “social credit” system that is using AI to quantify human behavior and interaction.
The danger in these systems isn’t their existence, because they are simply tools that can be used for good. The danger is their abuse.
A good manager can use a system like this to supercharge employees by focusing on their happiness – which can be measured – make sure they aren’t overworked, and highlight incredible customer service examples.
A bad manager can use a system like this to micromanage and create a “big brother” culture where it’s no longer enough to be on time and get your work done, but the nuances of your every word and emotion are monitored by AI.
Just like the power of Nuclear Energy, the innovation of AI can be used for both good and bad.
A very large cultural-generational shift has been happening for some time, driven by technology behaviors: the gradual abandonment of privacy concerns.
A range of factors contribute to this, including a greater level of trust in many corporate institutions vis a vis keeping their data secure and the very different media landscape that has evolved. Millennials and Gen Z are growing up doing thing like:
- Streaming video of themselves playing videogames in their room to the world
- Sharing the most intimate details of their lives on Twitter and Instagram in public
- Aspiring to join the “democratization of celebrity” that social media channels like YouTube provide
In a study I read years ago, the #1 career aspiration for tweens was to become a YouTube celebrity.
I myself have happily installed several "surveillance" devices in my home, and even purchased one from Google to build myself from scratch.
The long-running explosion of social media has driven the erosion of privacy concerns in heavy users, and I think at a macro-level, this is going to help accelerate the adoption and use of artificial intelligence, as AI requires data for fuel.
See more of our thoughts on the Future of Commerce in our webinar exploring the intersection of AI and Mixed Reality below.