Skip to main content Skip to navigation

NewsGPT: ChatGPT Integration for Robot-Reporter

Project Overview

The document explores the integration of generative AI, particularly OpenAI's GPT model, with robots like Pepper to improve interactions in various domains, including education. By enabling robots to conduct natural language conversations, understand user inquiries, summarize intricate information, and provide real-time updates, the framework enhances engagement and information delivery. In the educational context, this technology holds promise for personalized learning experiences, where AI can assist students in understanding complex subjects, answering questions, and providing tailored feedback. The findings suggest that while generative AI can significantly boost efficiency and accuracy in educational settings, it is essential to recognize the irreplaceable qualities of human educators, such as empathy and nuanced understanding. Ultimately, the document highlights the transformative potential of generative AI in education, noting that its applications can lead to more interactive, responsive, and effective learning environments while complementing traditional teaching methods.

Key Applications

Integration of GPT model with Pepper robot for news reporting

Context: The framework is designed for journalism, enabling robots to report news and engage with users in educational settings.

Implementation: The system uses voice input, speech-to-text transcription, context analysis, and text-to-speech synthesis to enable conversations.

Outcomes: Improved conversational capabilities of robots, faster news delivery, and enhanced user engagement.

Challenges: Limitations include the robot's struggle with context retention and response accuracy for broad queries.

Implementation Barriers

Technical Limitations

The robot struggles to retain context over the course of a conversation.

Proposed Solutions: Further development of memory and contextual awareness in AI systems.

Response Accuracy

Inaccurate or irrelevant responses for broad news queries.

Proposed Solutions: Improving the model's training data and algorithms to enhance comprehension and response generation.

Project Team

Abdelhadi Hireche

Researcher

Abdelkader Nasreddine Belkacem

Researcher

Sadia Jamil

Researcher

Chao Chen

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Abdelhadi Hireche, Abdelkader Nasreddine Belkacem, Sadia Jamil, Chao Chen

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies