AI and User Privacy
In my previous article UX Design for AI Product, I mentioned that “AI is smart, it make decisions based on what it knows about you. Sometime this freaks user out. It’s important for designers to think about how do we protect user’s data so user don’t feel that the AI tracks everything about them.” So I took some thoughts about this topic.
One day I went to this restaurant I used to go to, the waitress recognized me.
Waitress: “Hey, I remember you, you were here last week! Do you still want that Bento box you used to have?
It’s was that moment I feel very special. The waitress remembered me and she even knew about my preference.
Me: “Yes, I would like to have the bento box today.”
Waitress: “You still want hot tea, right?”
Me: “Yes that’s right! Wow you remember that, that’s very considerate!”
I felt I am a valued customer who just got the best services.
Waitress: “Hey actually, we have a Wagyu bento box today, it’s special of the day and I think you might like it.”
Me: “Oh really, I would love to try that!”
Here the waitress has successfully recommended a dish to me since she remembers what I like.
This reminds me some customer engagement strategies: such as the recommendation modules on the eCommerce website “Products to buy inspired by your browser history”, which is used to give user more choices and encourage them to buy more goodies; In a news app, “Hey this article might be of your interests since you shared a similar article on your news feeds”. The website remembers what you browsed and then recommend something based on your preference so you will read more similar contents. In the era of big data and artificial intelligence, we more or less capture information about our users, it’s key for designers to make sure users aware of what we are tracking and why is that. LinkedIn did a good job explaining to the users what information is being tracked, and why it’s being tracked.
This is the part that AI has done right. Just like that waitress, I don’t mind that she noted down my food preference, since these are relevant information to the dinning experience. In fact, I feel great about that she remembered these information, cause it helps the restaurant serving me better.
However, let’s imagine that the waitress is a robot. In order for a robot to remember me, to identify my face, remember my behavior and preference, the robot technology probably involves facial recognition and video recording of my personal images, aka, my privacy data. And you know what, the robot is very likely recording me without me noticing it. Thinking about my face and images are stored in the back end database, and it could be pulled out by some back end engineer, this kinda freaks me out. I actually worried about this scenario when I walk by my Nest camera at home everyday. ;-) I think it is very important for the robot company to make sure these privacy data is securely stored, and it can’t be pulled by any individual.
Then, I imagined this scenario.
Waitress: “Hey since you come to our restaurant almost every week, do you want to keep your credit card information on our file, so next time you can pay easily?”
Me: “Hmm, let me think about that. You know what, maybe not, but thank you for your offer.”
This is the time when I feel a little bit on the fence. Credit card information is a sensitive information to me, although I trust this waitress and I believe she will not do dishonest charges on my card, but I am still pretty conservative about this. For some user who shops on an eCommerce site a lot and have an account, they are probably OK about sharing their credit card information to the site since it will literally save them a lot of time and stream line the shopping experience. But not everybody like that idea. For the information that we are not sure if the users are willing to give us, it’s important to always ask them to grant the access. That’s why it’s important for the eCommrce site to ask “hey do you want me to remember your credit card info, so next time you don’t have to enter these information again.”
Then I took a bold imagination about another scenario:
Waitress: “Hey can I have you name and phone number and home address? I can call you when our restaurant have some special dishes!”
Me: “Oh no, I don’t think so. That’s my private information and I don’t want to share it to the public.”
This is the time the waitress has crossed the line. The waitress doesn’t need to know everything about me in order to serve me a good dish. The same rule applies to AI technology. For user privacy, we need to filter out the irrelevant information, and find out what are the relevant information that can actually improve user experience, and the users are OK for the AI to track. In this way, we can use AI to serve our customer better, while keep them feel safe about their privacy.
