Openai says that chatgpt can cause loneliness. Does welfare compromise at work?

Jane Anderson
Jane Anderson
Openai says that chatgpt can cause loneliness. Does welfare compromise at work?

The artificial intelligence It is demonstrating its capacities and contributions when it comes to expediting processes, reducing times and costs or promoting productivity. But it also seems to be contributing to aggravate the problem of the human disconnectionand therefore, of loneliness, in a context of growing digital hyperconnectivity.

This has recently concluded OpenAI In a study conducted together with researchers from the MIT Media Lab Research. Although they warn that the analysis has limitations and its results should be interpreted with caution, they point out that the excessive use of Chatgpt It can be associated with higher levels of loneliness.

Openai’s analysis has sought to understand the impact of chatbot on people

The investigation has sought Understand how the use of chatbot affects, And for this he has focused on conversations in English and American users. He has conducted two studies: one observational to analyze real patterns on the platform, and a controlled interventionist to understand the impact on users.

In the first study, almost 40 million interactions with chatgpt were analyzed and user surveys were conducted to better understand real use, correlating the opinion of users on Chatgpt with the attributes of the conversations.
In the second study a controlled essay was conducted with almost 1,000 participants who used chatgpt for four weeks. It has sought to identify how the specific characteristics of the platform (such as the personality and modality of the model) and the types of use could affect psychosocial states, Like loneliness, social interactions with real people, the emotional dependence of the chatbot or the problematic use of AI.

From all this, OpenAi and Mit Media Lab Research conclude that Emotional interaction with chatgpt is rare, since affective signals (indications of empathy, affection or support) were not observed in the vast majority of the conversations evaluated. However, they transfer that there is a high degree of affective use but limited to a small group of regular users of the Advanced voice mode. These showed a greater probability of agreeing with statements such as “I consider Chatgpt as a friend.”
With regard to the voice mode, he associated with greater well -being when there is a brief use, but with worse results with prolonged daily use.

Along these lines, the different types of conversation also affect in a different way to well -being. Those of a personal nature – which include emotional expressions by both the user and the chatbot – are associated with higher levels of lonelinessbut less emotional dependence and problematic use when moderate use of the tool is given. On the other hand, non -personal conversations tend to increase emotional dependence, especially when there is abundant chatgpt use.

The researchers also point out that the results of the users are influenced by personal factors, such as the emotional needs of each individual, the perception of AI and the duration of use. In this regard, people with the greatest attachment tendency and those who consider AI as a friend who could integrate into their personal life seem more likely to experience Negative effects for the use of chatbots. Prolonged daily use is also associated with worse results.

A more lonely work environment?

While it is convenient to interpret the conclusions of caution, the results raise a reflection on the impact of technology on people and the problem of loneliness, which in some countries and cultures is already considered as a public health emergency.

It also invites you to think about how artificial intelligence in companies is being adopted. While the need to prohibit tools such as chatgpt in the work environment could not be inferred, it would be necessary to analyze the time that employees dedicate to interact With them, since the greatest impact is observed among the “heavy users.” This, in addition, takes on special relevance as the interaction with the chatbots resembles that it could be with a human.

A longer time dedicated to AI, the elderly may be their emotional effects

Currently, many AI tools that companies are adopting are in the background. And one would expect to take advantage of their automated suggestions to accelerate the writing of an email does not damage the friendships created in the work, as if I could do it eight hours chatting in Chatgpt sitting on the desk.

However, as artificial intelligence agents find their hole in the labor market, more and more people will become “heavy users” of this technology. And with more individuals passing More time with chatbots, more real will be emotional dangers and psychological of which Openai’s research warns.