After two hours of interview, AI recreates human personality with 85% accuracy

A new technology developed by researchers in collaboration with the Google DeepMind is opening revolutionary paths for artificial intelligence and sociological research. The project uses AI to create digital replicas of the human personality, based on interviews lasting just two hours.

How does the technology work?

Innovation is supported by advanced language models (LLM), such as ChatGPTand is designed to conduct in-depth interviews. The system analyzes the interviewee’s answers and generates a digital simulation of the personality, capable of answering questions with consistency similar to the real person.

The process includes multiple steps:

1. Interview: AI asks targeted questions, listening and analyzing answers.
2. Processing: Data is interpreted to model personality.
3. Validation: the “artificial personality” is subjected to tests that compare its answers to those of the real interviewee.

In tests, the system demonstrated 85% accuracy in the correspondence between AI responses and those provided by real people. This result was obtained from training with a thousand volunteers, who participated in the interviews.

This technology aims to transform the way sociological research is conducted. Today, survey-based studies require a lot of time and resources for data collection and analysis. With the new approach, it would be possible to carry out broader and more economical studies, accelerating the process and expanding the understanding of social dynamics.

The researchers named the simulations “agents”setting them apart from common virtual assistants. These agents represent a new class of AI, with more advanced behaviors.

Digital replicas can be useful in several areas

Digital personality replicas offer intriguing possibilities in several areas, such as:

• Virtual assistants: improving personalization and efficiency in human-machine interactions.
• Robotics: optimizing interactions in home and professional environments.
• Education and training: simulating human behaviors for specific case studies or practices.

Ethics and privacy in focus

Despite technological advances, the research raises important ethical questions, such as the responsible use of personality replicas and the protection of the privacy of interviewees’ data. The creators emphasize that the The objective is not to replace human beings, but to complement sociological studies and improve interactions between humans and machines.

Next steps

Published on the arXiv platformthe study is a milestone in the application of AI to understand human behavior. It also reinforces the position of artificial intelligence as a powerful tool in researching and creating solutions to social and technological challenges.