1. What is this AI chatbot, and how does it work?
Our AI chatbot is an advanced search tool that helps community members quickly find relevant information. It uses Retrieval-Augmented Generation (RAG), a modern way of searching that retrieves documents from our community’s knowledgebase and presents answers in a conversational format. It also leverages a large language model (currently IBM’s Granite model, or OpenAI’s GPT-4o mini model) to enhance responses when appropriate.
2. What is RAG, and how is it different from traditional search?
RAG is just a smarter way to search. Traditional search engines return a list of documents based on keywords, leaving you to sift through them. RAG works by retrieving the most relevant documents and generating a summarized response while citing the sources. This means you get direct answers with references to the original content.
3. Does the chatbot create new knowledge?
No, the chatbot does not create new knowledge. It [primarily] retrieves and summarizes information from trusted sources within the community. [The chatbot can also answer general industry-related questions using the LLM’s training data—but only when no more relevant community-generated content is available.]
4. What sources does the chatbot use to generate answers?
The chatbot retrieves information from:
• The community’s knowledge base (e.g., past discussions, blogs, uploaded documents)
• Official resources of the organization
• [The LLM’s training data, but only if it is relevant and no better community or official content exists]
5. Is my past content being used without my permission?
No, your contributions remain protected. The AI does not store or reuse your past content beyond the specific response it generates in the moment. It functions just like a search engine that pulls up relevant posts but does not retain any information afterward.
6. How does this chatbot protect privacy and content security?
We use Azure OpenAI and IBM watsonx, which both ensures that every request is processed securely. The chatbot does not retain or learn from past interactions—each question is treated as a fresh search. Your contributions remain safe and private.
7. Will my contributions be credited if they are used in an answer?
Yes. Since the chatbot retrieves relevant documents, it can provide references to the original sources. This ensures transparency and allows members to see where the information is coming from.
8. Can I request that my past contributions not be used?
Since the chatbot only retrieves existing content like a search engine, we follow the same policies as our community terms and conditions.
9. What happens if the chatbot can’t answer my question?
If the chatbot is unable to provide a satisfactory answer, you can use our “Ask Your Fellow Humans” feature! Our community of experts in our field are all gathered together in this forum. The perfect place to ask the experts.
[10. Is this replacing human moderators or customer support?
No, the chatbot is meant to assist, not replace, human engagement. It helps answer common questions quickly, but members can still connect with experts and moderators for complex discussions.]
11. What happens if the chatbot gives an incorrect or misleading answer?
While RAG improves accuracy by pulling from real sources, mistakes can happen. If you notice incorrect information, you can report it, and we will review and refine the chatbot’s knowledgebase to improve future searches.
12. How can I trust that the AI is not biased?
The chatbot prioritizes “official content” from the organization, and trusted community-generated content over general AI-generated responses. This ensures that industry experts, professionals, and community members drive the accuracy and relevance of the information.
13. Can I see the sources used to generate an answer?
Yes! The chatbot provides references for the information it retrieves, so you can always check the original sources.
14. Why is this better than just using Google or ChatGPT?
Google searches the entire internet, which may include unreliable or irrelevant sources. ChatGPT is trained on those same sources Our chatbot is focused on content created by us and our community of experts, making it more precise and useful for members.
15. Will AI take over all community discussions?
Absolutely not! The chatbot is a tool for enhancing engagement, not replacing discussions. It helps surface useful information faster, so members can focus on deeper conversations and collaboration.
16. How can I provide feedback on the chatbot’s performance?
We welcome your feedback! You can report issues or suggestions within the chatbot interface or contact our support team to help refine the experience.