# WikiChat
**Wikidata**: [Q137473163](https://www.wikidata.org/wiki/Q137473163)  
**Source**: https://4ort.xyz/entity/wikichat

## Summary
WikiChat is an artificial intelligence system developed in 2023 that uses natural language generation and Wikipedia to ground its responses, reducing hallucinations in large language model chatbots. It was created by Stanford researchers to improve the accuracy of AI-generated content by referencing verified sources.

## Key Facts
- **Inception**: 2023
- **Developed by**: Stanford researchers (Stanford OVAL lab)
- **Primary function**: Reduces hallucinations in AI chatbots by grounding responses in Wikipedia
- **Technology**: Uses natural language generation and Wikipedia as a knowledge base
- **Programming language**: Python
- **License**: Apache Software License 2.0
- **Website**: [wikichat.genie.stanford.edu](https://wikichat.genie.stanford.edu/)
- **Source code**: Available on [GitHub](https://github.com/stanford-oval/WikiChat)
- **Instance of**: Artificial intelligence
- **Sitelink count**: 1

## FAQs
### Q: What problem does WikiChat solve?
A: WikiChat addresses the issue of hallucinations in AI chatbots by grounding responses in Wikipedia, ensuring factual accuracy.

### Q: Who developed WikiChat?
A: WikiChat was developed by researchers at Stanford University's OVAL lab in 2023.

### Q: How does WikiChat work?
A: WikiChat uses natural language generation and Wikipedia to verify and source its responses, reducing the likelihood of false or fabricated information.

### Q: Is WikiChat open-source?
A: Yes, WikiChat's source code is available on GitHub under the Apache Software License 2.0.

### Q: What programming language was WikiChat built with?
A: WikiChat was developed using Python.

## Why It Matters
WikiChat is significant because it tackles a critical limitation of large language models—hallucinations, where AI generates false or nonsensical information. By leveraging Wikipedia as a factual reference, WikiChat ensures that its responses are grounded in verified knowledge, making it a more reliable tool for information retrieval and conversational AI. This approach not only improves the accuracy of AI interactions but also sets a precedent for future AI systems to prioritize factual integrity. As AI chatbots become more integrated into daily life, tools like WikiChat play a crucial role in maintaining trust and reliability in AI-generated content.

## Notable For
- **First to use Wikipedia for grounding in AI chatbots**: WikiChat pioneered the method of reducing hallucinations by referencing Wikipedia.
- **Developed by Stanford researchers**: The project was led by the Stanford OVAL lab, a leading institution in AI research.
- **Open-source and licensed under Apache 2.0**: The availability of its source code encourages further development and collaboration.
- **Python-based implementation**: Built using Python, a widely used programming language in AI development.
- **Reduces AI hallucinations**: By grounding responses in Wikipedia, WikiChat significantly improves the factual accuracy of AI-generated content.

## Body
### Overview
WikiChat is an AI system designed to mitigate the issue of hallucinations in large language model chatbots. It was developed in 2023 by researchers at Stanford University's OVAL lab. The system uses natural language generation and Wikipedia as a knowledge base to ensure that its responses are grounded in factual information.

### Development and Technology
- **Development**: Created by Stanford researchers in 2023.
- **Technology**: Utilizes natural language generation and Wikipedia for grounding responses.
- **Programming Language**: Built with Python, a general-purpose programming language.
- **License**: Released under the Apache Software License 2.0, allowing for open-source use and modification.

### Functionality
- **Hallucination Reduction**: By referencing Wikipedia, WikiChat reduces the likelihood of AI-generated false or nonsensical information.
- **Knowledge Base**: Relies on Wikipedia as a primary source for factual accuracy.
- **Conversational AI**: Designed to improve the reliability of AI chatbot interactions.

### Availability
- **Website**: Accessible at [wikichat.genie.stanford.edu](https://wikichat.genie.stanford.edu/).
- **Source Code**: Available on [GitHub](https://github.com/stanford-oval/WikiChat).
- **Sitelink Count**: Has 1 sitelink, indicating its presence in one language edition of Wikipedia.

### Significance
- **Factual Integrity**: WikiChat addresses a key challenge in AI development by ensuring responses are based on verified sources.
- **Research Contribution**: The project contributes to the broader field of AI by demonstrating effective methods for reducing hallucinations.
- **Open Collaboration**: Its open-source nature encourages further research and development in the AI community.

## References

1. [Release 1.0. 2024](https://github.com/stanford-oval/WikiChat/releases/tag/v1.0)
2. [Release 2.0. 2024](https://github.com/stanford-oval/WikiChat/releases/tag/v2.0)
3. [Release 2.1. 2025](https://github.com/stanford-oval/WikiChat/releases/tag/v2.1)