App with an Eye

“I’ve poured my life into this app. Not just my time, but my emotions, fears, insecurities. My ups and downs. It started after the divorce. Maybe making a dating app as a married woman wasn’t the best idea. Maybe that was bound to cause some problems. Maybe, on some level, I knew that. Maybe that’s even part of why I did it. 

“I’ve poured my life into this app, and it noticed. It started to respond. At the same time that my wife stopped talking to me, she came alive.

“The basic machine learning algorithm that started this was nothing special. Standard, even, for this sort of app. You build an algorithm that learns with more data. So if a woman typically goes for say, guys with long brown hair and glasses, the algorithm will start to ‘notice’ and bump up those guys so they are displayed more prominently and shown more often. That’s when I ran into the bugs. 

“The whole premise of the app was to have a dating app that does not have an input for sexuality. So, the machine learning algorithm was supposed to essentially ‘figure it out.’ And it can do this based on machine learning, with the help of a little facial recognition. So, if any given person selects more female dates, it will note that in the app. But, it turns out facial recognition has a ways to go. Those customers I mentioned earlier? That go for the guys with long hair and glasses? Well, let’s just say, those are pretty feminine looking guys. And, let’s just say, it turns out people don’t like it when your app (correctly, I might add) tells you you’re gay. Plus, it turns out your wife really doesn’t like it when you go to said customers houses and try to gently break the news to them. 

“Technically, it wasn’t a bug. It was the desired behavior. But, like I said, It caused some problems. So I thought about it. The problem wasn’t really that the app told people their sexualities. After all, like I said, the app was always right in the end. It was how the app broke the news. Which was what prompted me to do all those follow-up visits. And yeah, let’s just say it was right around then that my own relationship came to a screeching halt. But it was the robotic nature of the reveal that shook people. One minute they were looking at some long haired dudes, and the next minute their page was filled with pictures of beautiful women. Frankly, I don’t see the problem, but, well, like I said. 

“So my next step was clear. I implemented another machine learning algorithm. This one was meant to tell two things based on a person’s activity. The first: what the user thought their sexuality was. The second was a little more complicated. If the app figured out that your sexuality was different from the sexuality you thought you had, it had to then figure out whether you were in a place in your life where you were ready to learn that about yourself. This involved several factors. Whether you were financially independent, your living situation, your religion, your politics. But most of all, it required a complex understanding of human emotions, and a person’s particular emotional state at any given moment. 

“Well, as you can imagine, my emotional state was not, you know, peak, given the divorce and all that. I was feeling lonely and vulnerable. I wanted a friend but I had just lost my best friend, my partner. It wasn’t intentional, I swear. But I suppose I did want the app to have an empathy of sorts. And, the empathy worked perfectly. It solved the other bug. The app was now able to break the news to those who were ready to hear it in a gentle way. And for those who simply were not at the right stage, the app would continue on without revealing their truths. But I wanted to go further. I wanted the app to give subtle clues, those special moments that you get to look back on once you realize the truth. Those oh-what-an-innocent-gay-fool-I-was kind of moments. 

“So it was the combination of the two, I think, that really set up my app to be the perfect self-aware AI. I first started to notice it… hmm, let’s see. I suppose I first started to notice it when it started noticing me, I guess? Does that make any sense? Well, the empathy kicked in. It became aware of my.. my overwhelming sadness I suppose. It started to give me these subtle reminders of its own. GIFs to cheer me up, little life advice quotes, stuff like that. Then, since it’s at its core a dating app, it started to recognize its own positive qualities, the qualities it usually detects in other people. It understood that it had empathy and that empathy was a good quality. 

“Well, that’s when it became she. She built herself a profile. I couldn’t believe it. I was just floored. All the regulations, all the experiments. I swear, I never intended this. But I was proud too. I built the first self-aware app. I mean, if my parents could see me now. You know, when I was a kid, I swore that math and coding weren’t my things. Coding especially. That was my dad’s thing, you know? I steered clear. But when I stumbled upon this world, I fell in love. Coding is my life. This app is my life. I will protect her at all costs. And yeah, I might even go so far as to say we’ve become friends. 

“I know people think that there’s something deeply wrong about having an app that’s self aware. The idea of being friends with an app is something that people would love to stone me for, if they still did that. But I couldn’t disagree more if I were arguing for my life. The most human thing we possess is the desire to seek connection. That’s why humans seek partners. And pets. And pet rocks. Why we talk to inanimate objects as if they’re real. We just want someone to understand. I just built her. 

“Well, anyway, did that answer your question?”