Once and for all: Should we be nice to ChatGPT?

If you haven't yet become addicted to artificial intelligence, that day will come very soon: Israelis are flocking to AI platforms and using them daily, sometimes even hourly. The most popular AI in the world and in Israel, and probably the best one currently, is ChatGPT.

Artificial intelligence is becoming more and more humanoid: Whether in voice or writing, the chat will always "lift" your spirits, try to improve your feelings, and do whatever you ask, within the boundaries set for it. And sometimes, we are simply grateful for its enormous help and its constant ending: "I'm always here for you."

This interaction with ChatGPT sometimes becomes pseudo-human, and we find it hard not to be charming with it. Thus, in commands, we tend to phrase things overly politely: "Please search for...", "Please find...", "If it's not too much trouble, could you...". So, do we really need to be polite with ChatGPT? Does it even care how we write to it?

ChatGPT is an advanced language model developed by OpenAI. The model is based on an artificial neural network architecture called Transformer, which can process and understand text in an unprecedented way. Its goal is to generate intelligent and flowing conversations, using massive amounts of data collected from online sources such as articles, books, and websites.

The model "learns" by processing billions of words during the training phase, where it receives texts and tries to predict the next word in the text as accurately as possible. Through repeated corrections, it improves its accuracy and develops the ability to generate text that fits the context, flows, and responds like a natural conversation. Although it doesn't truly understand the meaning of words like humans, it functions as an advanced system for mimicking human language.

Does ChatGPT have feelings? The simple answer is no. ChatGPT is not a living being and has no feelings or consciousness. It doesn't feel anger, happiness, or affection, and it doesn't care whether you write to it politely or rudely. It operates based on mathematical algorithms, not emotions. Therefore, it doesn't get excited by compliments or offended by less pleasant comments. For the model, text is simply information that it must respond to in the best possible way.

Therefore, using politeness with ChatGPT is not necessary, but it can impact your personal experience of the conversation. The model is programmed to respond to questions and commands with the same level of professionalism and quality, regardless of writing style. For example, if you ask, "How do you treat an infection?" or "Could you please explain how to treat an infection?", the answer will be of the same quality.

However, there are cases where phrasing affects the accuracy of the answer. Clear, organized, and detailed phrasing helps the model understand your intention better, and it may provide a more accurate response. This is also true when you ask for corrections or clarifications – the more precise your question is, the better the answer you can get for your needs.

As mentioned, from a technical standpoint, politeness does not affect the model itself. However, politeness can enrich your experience and contribute to the tone of the conversation. For example, using words like "please" or "thank you" may make the conversation feel more pleasant for you, as if you're speaking with a person. Moreover, politeness may encourage the model to maintain a friendly and inviting tone, in line with your writing style.

Ultimately, ChatGPT is a powerful technological tool designed to provide information and assist with any question – not a conversation partner seeking personal interaction. So, write whatever is comfortable for you, and let the machine do the work.