What type of questions can i ask
#What kind of questions can i ask?
north kernel
glossy jayBOT
Hi @north kernel. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
You can ask a wide variety of questions when using OpenAI models and tools, but the types of questions you can ask—and expect good answers for—depend on the context, the model, and the setup:
-
General Knowledge & Factual Questions: You can ask for facts, summaries, or explanations about a wide range of topics. For example, "Which athletes won the gold medal in curling at the 2022 Winter Olympics?" or "What are three compounds we should consider investigating to advance research into new antibiotics?" These are well-supported by OpenAI models, especially when relevant context is provided or retrieved using embeddings or retrieval-augmented generation (RAG) systems Question answering using embeddings-based search, Reasoning models: Prompt examples.
-
STEM and Research Questions: OpenAI's o-series models (like o4-mini) are particularly strong at supporting scientific and technical research questions, such as those related to STEM fields Reasoning models: Prompt examples.
-
Custom and Domain-Specific Questions: If you have set up a retrieval system (e.g., using embeddings, vector databases, or file search), you can ask questions that are answerable from your specific knowledge base or documents. For example, "How did COVID-19 affect the 2022 Winter Olympics?" or "What new capabilities will ChatGPT have as a result of the partnership between OpenAI and Schibsted Media Group?" Question answering using embeddings-based search, Doing RAG on PDFs using File Search.
-
Open-Ended, Subjective, or Creative Questions: You can ask for opinions, summaries, or even creative outputs (like poems), but the quality and appropriateness of the answer may depend on the model and the system's configuration. For example, "Which Olympic sport is the most entertaining?" or "Write a four-line poem about the elegance of the Shoebill Stork" Question answering using embeddings-based search.
-
Instructional or Task-Oriented Questions: You can ask the model to perform tasks, such as generating questions from a document, summarizing content, or extracting entities from a prompt Doing RAG on PDFs using File Search, RAG with a Graph database.
Limitations:
- If your question is outside the scope of the provided data or context, the model may respond with "I could not find an answer."
- For retrieval-augmented systems, questions should be answerable from the available documents or data.
- Some systems are configured to only answer questions on a specific topic (e.g., "You answer questions about the 2022 Winter Olympics.").
In summary, you can ask factual, research, domain-specific, creative, and task-oriented questions, but the best results come when your question matches the model's strengths and the available context or data.
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
north kernel
@glossy jay id like to know more about discord webhooks, news channnels and how to add them to my own small server
glossy jayBOT
Let me check the knowledge base for you again. One moment! :books: