Empathy Unleashed: AI Disclosure Drives User Acceptance

summary: Disclosure of personal information by artificial intelligence (AI) agents promotes user empathy and increases acceptance of AI technology. This study involved an online experiment in which a participant had a text-based chat with her AI agent.

Researchers found that when agents disclosed work-related personal information, it was associated with greater empathy from participants, but disclosure did not suppress empathy.

The AI ​​agent’s visual representation, whether human-like or robot-like, did not significantly affect its empathy level.

Important facts:

1. The study found that disclosing personal information, especially work-related information, by AI agents increases user empathy and improves acceptance of AI technology.
2. The visual representation of the AI ​​agent, whether human-like or robotic, did not significantly affect the level of empathy participants felt toward the AI.
3. This study involved an online text-based chat where an AI agent and participants participated in a scenario as if they were colleagues on their lunch break, suggesting that context and relational dynamics play a role in the outcome of human-AI interaction. It shows that you can fulfill

sauce: PLOS

A new study found that participants showed more empathy when an online anthropomorphic artificial intelligence (AI) agent appeared to disclose personal information about themselves while chatting with participants.

Takahiro Tsumura of the Graduate University for Advanced Studies (Tokyo) and Seiji Yamada of the National Institute of Informatics (Tokyo) have published these findings in an open access journal. pro swan May 10, 2023.

With the increasing use of AI in everyday life, there is growing interest in the factors that may contribute to the level of trust and acceptance that people feel toward AI agents.

Previous research suggests that people are more likely to accept man-made objects if the objects elicit empathy. For example, people may relate to cleaning robots, robots that mimic pets, and anthropomorphic chat tools that offer assistance on websites.

Previous research has also highlighted the importance of disclosing personal information in building relationships.

From these findings, Tsumura and Yamada hypothesized that self-disclosure by anthropomorphic AI agents might increase people’s empathy for those agents.

To test this idea, researchers conducted an online experiment in which participants engaged in a text-based chat with an online AI agent visually represented by a human-like illustration or an anthropomorphic robot illustration. rice field.

The chat included a scenario where the participant and the agent were co-workers at the agent’s office during their lunch break. In each conversation, the agents appeared to disclose personal information that was highly work-relevant, less relevant about their hobbies, or not at all.

The final analysis included data from 918 participants who rated their empathy for AI agents using a standard empathy questionnaire.

Researchers found that highly job-relevant self-disclosures from AI agents were associated with greater empathy from participants compared to less relevant self-disclosures. Lack was associated with repressed empathy.

The agent’s appearance as a human or anthropomorphic robot was not significantly associated with empathy level.

These findings suggest that self-disclosure by AI agents may actually elicit empathy from humans, which may aid future development of AI tools.

The authors add: Our research will change the negative image of artifacts used in society and contribute to future social relationships between humans and anthropomorphic agents. ”

Funding: This work was partially supported by JST, CREST (JPMJCR21D4), Japan. This work was also supported by JST, Establishment of University Fellowships for the Creation of Science and Technology Innovation, grant number JPMJFS2136. Funders had no role in study design, data collection and analysis, publication decisions, or manuscript preparation.

About this Artificial Intelligence Research News

author: Hana Abdullah
sauce; PLOS
contact: Hannah Abdalla – PLOS
image: Image credited to Neuroscience News

Original research: open access.
“Effects of agent self-disclosure on human empathy” by Takahiro Tsumura et al. pro swan

overview

Effects of agent self-disclosure on human empathy

As AI technology advances, social acceptance of AI agents, including intelligent virtual agents and robots, becomes even more important for more applications of AI in human society. One of his ways to improve the relationship between humans and anthropomorphic agents is to make humans empathize with agents.

Empathy makes humans behave positively and kindly toward agents, making them more receptive. This study focuses on agent-to-human self-disclosure to enhance human empathy for anthropomorphic agents.

We ly investigate the possibility that self-disclosure from agents promotes human empathy. We hypothesize, ly analyze, and discuss the conditions under which humans are more empathetic to agents.

Experiments were conducted on agent appearance (human, robot), self-disclosure (relevant self-disclosure, irrelevant self-disclosure, no self-disclosure), and empathy before and after video stimulation. Analysis of variance (ANOVA) was performed using data from 918 participants.

We found that appearance factors had no main effect, and that self-disclosure, which was more relevant to the scenario used, promoted more human empathy, with a statistically significant difference. I also found that self-disclosure does not suppress empathy.

These results support the hypothesis. This study reveals that self-disclosure represents a key characteristic of anthropomorphic agents that help humans accept them.