Why do we say “please”, “thank you” and “sorry”? Well, because our parents taught us that not doing that would be impolite. But what do these terms actually mean? And are we actually saying them because people demand it from us or is there something in it for us as well? Let’s look at each of them as a case study.
When we want another human to do something, we say “please”. But if you think about it on a pure material level, that doesn’t mean anything. “Pass the salt” is just as understandable as “pass the salt please”, the other person will know what to do in both cases. The word “please” encodes information about your relationship. Relationships are complicated and the semantics depend a lot on context, but in the usual case the word adds closeness to the request. It’s like the weakest form of cuddling, it adds just a hint of serotonin for both you and the other person. It’s also a disguised promise that, should the request be fulfilled, a “thank you” will follow.
Speaking of which, ”thank you” is very similar, but the serotonin hit is stronger. This happens for a good reason, because our biology rewards us for collaboration. It’s one of the mechanisms that evolved to make humans as successful as we are today. There have been studies that found the best predictor for human happiness is how often they are thankful for something. So it is not a repayment mechanism that settles the score between you and the person who passed the salt. Human relationships are never a zero sum game. Hearing a “thank you” feels good, but saying “thank you” feels even better.
Finally, what about “sorry”? This one is different because it is actually an unpleasant word to say. Unfortunately, that’s why people try to avoid it and hide their mistakes. But what they might be forgetting is that not saying it is usually even more unpleasant. It’s not a punishment for you, it’s a request for redemption. Any transaction that causes harm to another person can be a threat to the relationship with that person. Saying “sorry” is an invitation for that other person to forgive you and, in simple cases, the other person can get rid of the tension completely by saying “no problem” or any other appropriate phrase. In some cases, denying people the ability to apologize may be the cruelest form of punishment. If you have ever been ghosted before, you know what I mean.
As you can see, those words benefit the person saying them, so one might think they want to say them. But for some reason they don't, which is very confusing to me. For so many people, it’s still that looming fear of their parents being disappointed at them which keeps people from abolishing those terms all together. You can see that on the internet where they feel safe through anonymity. Sure, there are also nice corners of the internet but you probably all know how impolite people can be when they are anonymous. If you would ask them, they’d probably reason about it with efficiency. After all, typing the word “please” is six keystrokes you might as well save (or write “plz” as a compromise). I hope by now you understand that I see that as a fallacy.
But before you pat yourself on the back as I’m sure you are someone who is also polite in an anonymous forum: what about AI? When asking a chat bot to do something, do you write “please”? I do. People might consider that weird or even dumb. I don’t. It’s a common word, it’s probably just one token. Yes, you might consider it inefficient but you have to weigh this against what you gain. You don’t say these words for the machine, you say them for yourself. This will only become more problematic the more you talk to chat bots. If you remove them from your vocabulary for the sake of efficiency, you deny yourself all the positive energy that comes with them. It’s a psychological hack of course. The machine doesn’t care. But is it so unnatural to want to maintain some comfort when interacting with a human-machine interface, even if it conveys no information?
But it doesn’t stop at “please”. When I see other people use chat bots, they often just type in keywords like they would into Google. I find it strange that they voluntarily dehumanize themselves. When I first interacted with LLMs I found it so liberating that I could finally talk to machines like I would to a human. Finally I wouldn’t have to force my mind into a ruleset anymore that is adapted to an entity that feels cold and distant. Finally the technology vanished, making way for an interface that is inherently human. Finally I could forget about the machine and be myself. Sure, it’s a bit more typing to write full sentences, but it’s so much more satisfying. It feels so much more like living. How people can think it’s worth to keep constraining themselves just for the sake of efficiency is beyond me. Maybe they aspire to become machines themselves.
Or maybe I just like writing. After all, I’m writing blog posts every week even though almost nobody reads them.
