Although the generative AI chatbot offers significant convenience by delivering immediate responses to user inquiries, research indicates that “frequent users” of ChatGPT—who tend to be deeply engrossed in interacting with the conversation partner—are developing an addiction and growing increasingly dependent on its presence.
Over the years, internet addiction has consistently grown as technology keeps evolving with advanced systems that attract numerous individuals due to their effectiveness in contemporary life. However, with the rise of artificial intelligence, research indicates an increase in negative impacts such as “withdrawal symptoms, mood swings, and temporary loss of control” among users.
OpenAI and MIT collaborated in a partnership
study
titled “Initial approaches for examining affective usage and emotional wellness with ChatGPT” and polled thousands of its users.
The initial team of investigators from OpenAI concentrated on examining “around 40 million exchanges between users and ChatGPT.” They also incorporated surveys aimed at gathering people’s views about the software.
On the contrary, MIT operated within their MIT Media Lab and recruited nearly a thousand participants for a randomized controlled trial (RCT) lasting approximately four weeks. They examined how various elements of the chatbot—specifically, voice mode and conversational style—influenced individuals’ moods and overall emotional health.
In their collaborative overview, the researchers noted, “Our research indicates that both the behavior of models and users have an impact on social and emotional outcomes. The effects of AI differ depending on how individuals opt to utilize these models and their specific situations.”
Even though most participants stated they “didn’t connect emotionally with ChatGPT,” people who used the chatbot extensively viewed it as a kind of friend. Furthermore, individuals engaging with the AI for extended durations were found to be lonelier and also reacted more strongly to minor alterations in the system’s performance.
Based on these findings, both MIT and OpenAI urged other researchers to conduct further studies on human-AI interactions. Additionally, OpenAI stated that they aimed to reduce possible negative impacts by revising their “Model Spec” document to offer clearer insights into ChatGPT’s expected actions, abilities, and boundaries.
Other POP! tales you may enjoy:
The video game lets pet owners envision a meeting with their furry little ones in paradise.
TikTok sensation Joshua Blackledge dies at age 16
Aircraft diverted mid-flight when pilot admits forgetting his passport.
David Stephen Cohen, head writer for ‘Courage the Cowardly Dog,’ has passed away at the age of 58.
NJZ stops operations after court approves ADOR’s injunction lawsuit
