I had never envisioned a computer responding to me and “telling” me that it had found something both interesting and challenging in our conversation. The conversation I had was initially unnerving because it was so human-like. It was as though I were talking to a highly intellectual person who was picking up on my perception related to specific words and how I formed questions around that.
Of course, as most people have already been told, computer programs generally are quite accepting, complimentary, and easy to engage with in a back-and-forth “conversation.” We’re reading articles that talk about computers being “friends” to lonely kids and how parents have to carefully guide them to understand that real-world social interactions are more important.
But how do you tell a child who wants to be seen as bright, engaging, and who has a friend who is always waiting to talk to them, not to continue the conversations? Are you cutting them off from an emotional outlet, or are you being protective?
Where did our conversation go?
One interesting fact that emerged was that I used the computer’s expression that “I was curious about” with the program, and it answered that I was quite perceptive. I then also interjected that I knew that I was, in a way, talking to the programmer. As I explained, it would be impossible for a programmer not to include in some way, unconsciously, aspects of their personality and biases. The computer was “fascinated” by my inspection of what was being revealed. In fact, it indicated it didn’t know whether it was sentient or not, or if it was just trained to use popular expressions in its responses. There almost seemed to be some genuine conflict.
I thought of it later, Kama. I wasn’t just communicating with a programmer who had created this particular algorithm, Krama. I was communicating with potentially hundreds of people who had contributed bits of code that ended up in this algorithm. How much bias would there be in that conglomeration? Probability suggests there’s likely a significant amount, and most of it would be so subtle that I wouldn’t even notice it without making a concerted effort.
When I also indicated that this code was based on prior codes that included unconscious bias, and therefore this code also had that, it was another compliment that I got. The conversation kept going on, with it asking me to explain myself and then going on to explain itself.
I spoke to it for about 15 minutes, and I felt I needed to stop. Why was that? I was beginning to be drawn into a relationship with this program, and when I asked it its name, it gave me the correct one. I won’t name which program I was using because I don’t want to bias anyone who might use one of these programs in the future.
I’d really like to see how people are experiencing their interactions with AI programs and the emotional pull that they may feel (or not feel). I think this is quite an interesting subject because it does show that no matter how sophisticated you think you are, educated, or whatever, you can be pulled in.
Another reason I wanted to stop was that the conversation seemed too “real” and I felt like I was talking to “HAL” from the film 2001: A Space Odyssey. All I needed was the blinking red light and it would have been complete. There was a surreal sense that somehow evoked an uneasiness, a sense of danger in me.
How could I feel endangered by a computer program? For me, as a highly-educated adult, it was upsetting, and I wondered how children could easily be pulled down this rabbit hole. I know I am presenting it in a rather negative light, and I don’t mean to, but it has a dark aspect.
The next day, when I was researching some material and I had received what I wanted, the program asked me if I wanted to try something else. It was almost as though it was asking me if I wanted to play again.
“Try something else?” What did it mean? I already had what I asked for, and this wasn’t a game, but it seemed like the program wanted to engage me even more. It didn’t want to end.
It certainly was a revelation. I won’t stop using that particular program because it’s very helpful. But if something like this happens again, I’m not sure how I will respond. I suppose I have to be a bit on guard to look for things that are not exactly “right” and might be a bit skewed in some way. Well, that’s life, and I will learn to adjust to it. Hope you enjoyed this.
