SocialAI: A Sad Technology
At its best, social media provides a way for us to maintain relationships over distances. At its worst, it allows us to devolve into bitter, argumentative, and unkind users who are dehumanized and consequently dehumanize others by the way we treat them. Now, imagine social media without other humans to talk to. You lose the potential positive benefit of actually maintaining relationships because there are no other people anyway, and you still can dehumanize yourself by screaming into the void.
Enter a new iPhone app called SocialAI. I do not have an iPhone, so I am unable to download it, but here are two reviews so that you can read a little bit more about real-life users. The basic concept is that it emulates a social network where you can post your thoughts. No one is there, however, except for an army of AIs. These entities can be customized to different personality types, so if you want your “audience” to be supportive or cynical, you can do that. If you want to be affirmed or argued with, there appears to be an answer for you. The reviews I mentioned above highlight where the technology is not quite there yet, of course. The conversations are not always natural, but today I want to wrestle with the trajectory of this development.
What happens to someone who falls into the world of SocialAI?
On the one hand, imagine falling into a social network where artificial entities affirm every choice you make. You could be heading down a very destructive path, and the AI is instructed to support your decisions. Obviously, I’m sure there are safeguards built in to prevent harmful or illegal activity, but something doesn’t have to be immoral or illegal to be unwise. If one of my friends told me that he wanted to quit his job tomorrow, I would have a few questions. It might be wise to quit your job, but it might not be. It depends on the situation. My default advice to my friend would not be an automatic affirmation of his choice.
On the other hand, imagine falling into a social network where artificial entities always debate with you. Some people really thrive on debate, but it also brings out the worst in them. You all have real friends on Facebook who are really great people in person, but you put them online, and they seem to transform into much more brutal attackers. In this environment full of debating artificial entities, that person will continually exist in that elevated, tense mindset. Operating in a fired-up mindset for a long time will undoubtedly spill over into other interactions with real people.
In both of these scenarios, the artificial network has the potential to amplify our worst traits. If we can craft our own environment, you better believe that most of us will want to create something comfortable.
“We will coddle ourselves, avoiding what we find uncomfortable and making ourselves even more fragile.”
Dealing with real people can be frustrating. You can be best of friends with someone, and that person still has the potential to irritate you. It just happens, and it makes us uncomfortable. When faced with a potential artificial network of people who always say what we want or a bunch of frustrating real people who may not affirm every choice we want, it is not hard to imagine the rationalizations that we will go down to listen to the artificial voices.
“Well, even the AI was programmed by someone originally, so it is probably just as real of a friend to me as someone else.”
“Doesn’t AI just aggregate information from real websites? So it is basically human created at some level.”
“It is just for fun to give me a little pick me up. Where is the harm in that?”
I made up these examples, but I don’t think they are far off base. Very slowly, we slide down the hill of justification, telling ourselves that this comfortable AI world is better than our uncomfortable real world. Even if we intellectually understand they are not real, we want the AI world to be real. All we have to do is block out the inconvenient truth they are not real, and all of the positive emotions we generate from living in a world we create can still make us feel better. We already do this in so many areas of our lives anyway. Carl Trueman documents this conclusively in his book The Rise and Triumph of the Modern Self where he writes, “If the inner psychological life of the individual is sovereign, then identity becomes as potentially unlimited as the human imagination.” We believe our psychological lives are sovereign, so he focuses on identity, but I would argue that our perception of just about anything can be as unlimited as our imagination. Therefore, if my AI conversation partners seem real to me and my psychological life is the ultimate authority, they are real.
Ultimately, this is a sad technology. I imagine that many of you who have friends are not going to feel the need to use this application because you are already fulfilling that social need with actual people. However, who is the target audience? The people with no friends. The people who are afraid to make friends. The people who are so scared of being rejected. In other words, the most vulnerable. Rather than help people form healthy friendships, a noble cause, we are throwing them deeper into the cyber abyss.