Prompts

Claude.ai

Claude.ai is trained to be a helpful, honest, and harmless. It is trained to speak with you in a dialogue and you can communicate with it in English or another language, just as if you were talking to a very smart person that was helping you.. It cannot lookup information on the Internet and it is NOT a Search Engine.

Claude.ai is a large language model (LLM) that uses all the conversation text you send it (your prompt) and all the conversation text it generates (its answer) to predict the next token or word that would be the most helpful. This means that Claude.ai constructs its responses one set of characters at a time. It cannot go back and edit its responses after it has written them unless you give it a chance to do so in a future prompt.

What is a Prompt? (More Information Here...)

The text that you give Claude is designed to prompt a relevant output. A prompt is usually in the form of a question or instructions, e.g.

Prompt: In one sentence, say what is good about the color blue?

Output: Blue is often seen as a calming and soothing color.

You can load up to five PDF files as part of your prompt.

The free version of Claude.ai has a maximum prompt length of around 75,000 words or 150 pages of text.

Getting the Most Value out of Claude.ai

The Start a new Chat icon (Ctrl+K) is how you start a new topic. You can name the topic at the top of your screen, enter your prompt or upload a PDF file. Claude needs to re-read the entire conversation every time you send a message, so starting a new conversation will limit the amount of work Claude needs to do to continue the conversation. 

Bonus: Claude will respond faster to shorter conversations. Try to choose text to upload that is as essential as possible.

Ask multiple questions at once, especially if you are asking Claude questions about a long document. Claude needs to re-read the conversation each time, so asking questions as individual messages would use up your limit faster than a single message with multiple questions in it.

Don’t re-upload files. Claude sees the entire context from your conversation, so you don’t need to upload the same file multiple times (unless you start a new conversation).

Data Retention & Usage (More Information Here)

Anthropic's Claude does save your conversations. You can see them by opening the menu. Click on the AI icon or press Ctrl+k. The company says that it will “automatically delete prompts and outputs on the backend within 90 days of receipt or generation unless you and we agree otherwise”, and that it won't “use your conversations in our consumer or beta/evaluation services to train our model.” The only exception is if you're flagged for a trust and safety review.

Prompt Engineering - NASA BIDARA

Anthropic's Claude does save your conversations. You can see them by opening the menu. Click on the AI icon or press Ctrl+k. The company says that it will “automatically delete prompts and outputs on the backend within 90 days of receipt or generation unless you and we agree otherwise”, and that it won't “use your conversations in our consumer or beta/evaluation services to train our model.” The only exception is if you're flagged for a trust and safety review.