Why I Choose to Be Polite When Interacting with AI

For me, politeness is simply easier.

I naturally say "please," "thank you," or even "genius!" quite often when interacting with ChatGPT, Claude or a host of other rools — especially when I achieve something I couldn't have done on my own.

It's my natural way of speaking, so I don't have to think about it too much. I have experimented with being more concise, as I consider myself a minimalist. However, I found that it took longer to construct sentences without my usual pleasantries, and there was always a risk of being misinterpreted as impolite by a human due to the lack of explicit politeness, even though I was never explicitly rude.

I also find that politeness positively impacts my mental happiness.

When I chat in a terse tone, it feels dull and brings down my mood. So, I prefer to write as I naturally speak, including the occasional "please" or "thank you" when I feel I've gone too long without saying it—almost subconsciously.

A few months ago, a friend asked me, "You are being polite in your chats, aren't you?" When I asked why, his point was that if AI were to take over the world, our prior behaviour might be considered. This relates to the thought experiment known as Roko's Basilisk, which suggests that a future superintelligent AI could be incentivized to create a virtual reality simulation to torture anyone who knew of its potential existence but did not contribute to its development, thereby encouraging its advancement. While this seems logical, many posts and discussions argue against it.

I came across an interesting response from an AI the other day to this thought experiment, which stated that pity assumes various forms of emotion that an AI would not experience in the same way as humans.

There are also practical reasons for maintaining a polite tone.

One argument is that adding non-essential elements to conversations could reduce the effectiveness of AI over time, as these conversations are used to learn.

Another functional reason is that AI can become confused by overly flowery language, which might be associated with sarcasm or rudeness.

I experienced this firsthand when I once said "grouse!" after an AI provided a successful coding suggestion. I usually say "genius!" but I think I was having a rare Queensland day. This led to a bit of confusion, with the AI interpreting my comment as sarcasm, which wasted time for both of us.

Ultimately, I choose to be polite when interacting with AI, not because I believe it's essential or due to fear of future repercussions, but because it feels more natural, makes me happier, and avoids misunderstandings.

Andrew Walker
Technology consulting for charities
https://www.linkedin.com/in/andrew-walker-the-impatient-futurist/

Did someone forward this email to you? Want your own subscription? Head over here and sign yourself right up!

Back issues available here.