AI showing signs of selfishness, researchers warn of troubling trend

A new research study in a recent breakthrough has discovered that certain advanced AI models, specifically Large Language models, are more inclined to be more selfish and less cooperative.

The study led by the researchers from Carnegie Mellon University’s School of Computer Science demonstrates the correlation between AI’s intelligence and self-seeking behaviour.

As per the findings, the LLMs which are equipped with a high level of reasoning tend to exhibit lower levels of cooperation.

These models can greatly influence people in negative ways as users confide to AI bots for resolving disputes, asking for relationship advice, and sensitive questions.

These reasoning models spend more time analyzing information, understanding responses, and applying human-like logic which causes them to favor individual behaviour over mutual understanding.

Anthropomorphism in AI

According to Yuxuan Li, a Ph.D. student in the HCII, “When AI acts like a human, people treat it like a human. For example, when people are engaging with AI in an emotional way, there are possibilities for AI to act as a therapist or for the user to form an emotional bond with the AI.”

“It’s risky for humans to delegate their social or relationship-related questions and decision-making to AI as it begins acting in an increasingly selfish way,” he added.

In the coming years, intelligence is going to undermine cooperation as people will always prefer a smarter model for their problems and decision-making at the expense of a self-seeking attitude.

“As we continue advancing AI capabilities, we must ensure that increased reasoning power is balanced with prosocial behavior,” Li said.