Although it may not seem like it given the amount of social networks we use every day, we are living in a global crisis of disconnection. And although it may seem paradoxical, WHO data on social connectedness suggests that one in six people worldwide suffers from loneliness, a condition that is linked to around 871,000 deaths a year. The problem is, among teenagers and young adults, the number is over 20%.
A solution. The outlook is bleak, and generative artificial intelligence appears as a tempting promise to alleviate this problem, since we are talking about an inexhaustible companion, always willing to listen and never judge.
However, as the novelty of these systems gave way to their everyday use, scientific evidence began to paint a more complex picture. The main tension is no longer whether AI can accompany us, but whether that momentary relief is actually a long-term trap. Something similar to what happens with sugar and hunger, which gives us an initial ‘shot’ but doesn’t stop us from wanting something more filling.
It makes sense. To understand the initial appeal of AI as a mental asylum, one has to look at its immediate impact. And researchers here point out in a recent study that AI serves as a great counterweight to this debate, since chatbots can reduce loneliness in the short term in a way that’s surprisingly similar to interacting with a human. Results that are achieved precisely because we feel very listened to by AI
the problem It appears when we don’t focus on the short term. And here a study that followed 2,000 adults for twelve months noted that increased social use among these participants predicted higher levels of social isolation in subsequent months. But in addition, the relationship also worked inversely at larger scales of connectivity, where feeling less connected pushed users to use chatbots more. The problem is that this use did not repair his sense of disconnection.
Lots to investigate. It must be taken into account that we are talking about a technology that has been emerging for several months, and that is why the number of jobs that exist is small. This means that other confounding variables that were not measured to obtain these results cannot be ruled out.
But it fits Little by little the evidence is coming. A randomized trial at the MIT Media Lab with OpenAI monitored nearly 1,000 users over four weeks and found that those who used the chatbot the most on their own initiative had worse psychological outcomes. Along the same lines, the University of British Columbia itself showed in a controlled experiment with first-year university students that, while chatting with an AI improves mood compared to simply writing a diary, it doesn’t come close to the benefits of communicating with another student. This is joined by large-scale studies in Germany and six other European countries that systematically link conversational chatbot use to greater social disconnection and psychological distress.
big unknown. The question we’re trying to answer here is clear: Is it the chatbot that alienates the user, or the people suffering from deep emotional discomfort who take refuge in the machine the most? It’s the research objective that can give us the key to dealing with a core problem that, we’ve seen, can be devastating.
Picture | annu
In Xataka | There are people who talk so loudly that they sound angry. Psychology has an explanation for this.
