The art of conversing with artificial intelligence

Many of those who have used GPT Chat They discover themselves including expressions such as “thank you very much” or “please” in their requests. Isn't it funny that we feel the need to be cordial and grateful to a machine? Isn't it ironic that sometimes we are kinder to an algorithm than to our siblings, partners, or neighbors? It turns out that this unexpected surge of friendliness reflects a surprising aspect of a conversation with an AI: a surprising feeling of intimacy.

The interaction with an interlocutor who seems human, but is not, usually clear tensions that disturb conversations with other people: the concern about appearing intelligent or cultured, the fear of ridicule or asking something stupid, the vanity of seeking to seduce others with our arguments. With AI, we converse with curiosity and a desire to experiment, and in some sense, we may find it easier to converse with a machine than with some people. We achieved turn the chat into a space conducive to abandoning shyness and recovering the best of our curiosity. We use conversations as a space for discovery and not confrontation.

An AI enables a conversation with an entity that does not judge us. We do not usually have trivial conversations with the tool and an exchange with the chat like those that occur in elevators, in which we talk about the weather or other irrelevant topics, would be almost unthinkable. Nobody opens an LLM to tell you that it is raining today or that it is hot. On the contrary, we arrive at the prompt eager to be given information about something we do not know and that we are seeking to learn. That is the essence of Socratic thought.

A dialogue of this type can also have a therapeutic effect. We live in a world where loneliness has become an epidemic. Paradoxically, there have never been so many people who feel alone as now, the era of hypercommunication. A study carried out in 2015 showed that loneliness has negative effects on our health comparable to smoking, obesity or addictions.

Can a machine work as an antidote to this modern disorder? We write this without ignoring how fragile and delicate this issue is. Clearly, it seems better to find suitable interlocutors in a partner, friends, family. In other people. The idea is not to replace these types of relationships. But perhaps we can stop feeling ashamed or rejected and ask ourselves why, since the origins of the very simple Eliza, conversation with programs that seem real sometimes fascinates and relaxes. For whatever reason, an AI can become one of the many tools that helps us deploy our own discourse and explore avenues, possibilities and questions that we find difficult to put on the table.

In 1563, the humanist Michel de Montaigne He locked himself in his castle to talk to himself about the most varied topics. Thus he invented the literary genre of the essay (in the etymological sense of “trying” or “trying” to think about ideas). One of them, perhaps the most famous, is his essay on the art of conversationwhich he conceives in complete solitude, in a conversation with himself. It is surprising that this text formulated in the 16th century continues to give us such accurate instructions on how to get the best out of a conversation.it doesn't matter if it is with another person or with an Artificial Intelligence:

• Find the proper order of our ideas, and carefully review our arguments.

• Embrace those who contradict us.

• Do not speak to convince, but to enjoy. Appreciate the exercise of reasoning.

• Doubt yourself and remember that we can always be wrong.

• Use conversation as a vital space to judge our own ideas.

• Keep critical thinking alive.

• Avoid prejudices, differentiating concrete examples from generalizations.

• Reflect on what we have learned from the other in the conversation.

We have seen many of these principles appear spontaneously in the conversations we usually have with Chat GPT. We are friendly, curious, we approach the conversation with the intention of discovering and not judging, we do not usually use the chat to impose our ideas, and we have a lively critical sense during the conversation. This may seem auspicious but it is advisable to be attentive to the habits that will be established, as the example of Twitter (today called X). In the beginning, this social network was not a toxic space, but rather a huge forum in which nerds, curious people and ingenious people gathered willing to share ideas. in 140 characters. However, over the years it has become one of the central stages of polarization of the time, and a space where aggression and disqualification proliferate to the detriment of constructive exchanges.

Are we moving towards a use of GPT Chat that offers us productive conversations, that helps us solve big problems and that allows us to find a middle ground between antagonistic positions? The experience of the deterioration of Twitter can help us see to what extent we have to be attentive to the use we give it.. Do we want to start a war with the machines and show that ChatGPT is stupid or do we want to keep it as a place that empowers our creations and gives us a chance to improve our thought process from a friendly place?