How to use ChatGPT
Generative AI has been a hot topic for well over a year, but there are plenty of us who are still learning about different AI tools and how to use them. If you’ve been interested in learning about ChatGPT but don’t know where to start, this guide is for you. We’ve put together a guide on how to use ChatGPT to support your nonprofit and save time. This guide will cover; the basics of how to use ChatGPT, best practices, how nonprofits can use ChatGPT, and a glossary of terms. People love using acronyms when it comes to AI, so we’ll do our best to demystify the technology and make it as simple as possible.
We won’t be getting into the ethics of using AI; that’s a topic that deserves its own significant consideration. We’ll leave it up to you and your team to determine if using the tool makes sense for your nonprofit.
You can also check out our guide on how to use Microsoft’s Bing AI Chatbot. Spoiler alert, we aren’t huge fans of the tool.
How to use ChatGPT
Navigate to chat.openai.com.
Select Log in if you already have an account. If you don’t, select Sign up and create an account. You’ll need to select an email address and password.
Use the text field to enter your prompt. Be sure to experiment with the type of prompt you’re requesting. For example, you can include the number of ideas that ChatGPT generates, ask it to use a certain writing style, or make your request extremely specific to your situation. Sometimes we’ll ask ChatGPT to give us 10 different ideas, or rewrite something in a more conversational style.
Best Practices
Verify that the information ChatGPT has produced is correct. While the technology is advanced, ChatGPT does occasionally give incorrect information. This is sometimes referred to as a “hallucination”. ChatGPT and other tools will confidently, and convincingly provide information that’s totally incorrect. A human should always review the content that ChatGPT uses for accuracy, especially before publishing the content online. Be aware that ChatGPT will make things up or include data that is untrue, and adapt accordingly.
Do not directly copy from ChatGPT onto your website, or any online communication. This is important for a number of reasons: the first being that AI generated text still looks and feels slightly unnatural and will not be a replacement for human written content. It would have been really easy for us to prompt ChatGPT to write this (or any other) blog post. It would also look and feel pretty weird. People claiming that ChatGPT or any other chat based AI tool are going to completely remove humans from manual writing tasks are overselling the technology.
Google and other search engines are able to identify AI generated text with amazing accuracy, and categorize it as spam. Having large blocks of AI generated text on your website can ruin your SEO (search engine optimization) and have a negative impact on your website’s online reputation.
Uses for Nonprofits
Generate ideas for content. Asking ChatGPT to give you a number of ideas for a blog or social media post, web page, or event announcement is a great way to get the ball rolling and start working through ideas that might suit your need. I hope you’re noticing a theme in this guide: ChatGPT should be used to supplement, not outright replace your existing processes.
Search for grants and foundations. ChatGPT can be used to search for grants and foundations. The best part is that you can include your nonprofit’s mission and programs and receive custom recommended grants. Be aware that ChatGPT has limited data from after 2021. Write a prompt like the following, and just replace the text in brackets with your own nonprofit’s information: “list grants that would be a good fit for the following nonprofit: a [region] nonprofit that [your nonprofit’s mission and programs go here]. Include urls to the organizations offering the grant”.
Summarize topics. If you’ve written a 1000 word response and need to narrow it down to 500 words (to satisfy a grant application, for example) copy the entire response into ChatGPT and prompt it to summarize the text to the desired word count. You can also do this to quickly summarize information that others have written for your own ease of reading.
Elaborate on topics. This is the inverse of the above suggestion. Let’s say that same grant application wants you to answer a prompt in 1000 words, and you’ve already covered this in 200 words. You can prompt ChatGPT to expound on your text and bring the word count up to the desired amount. Be sure to check all AI generated text for accuracy before publishing anywhere.
Generate first drafts and outlines. ChatGPT can be used to generate outlines for speeches, articles, blog posts, and most written content. While we don’t recommend copying large amounts of text from ChatGPT or any other AI writing tool, they often do a great job of generating ideas for content that are in a structure and format that can be easily improved upon.
Create formulas for Microsoft Excel and Google Sheets. ChatGPT can be prompted to create formulas for spreadsheets using plain english. For example, if you wanted a formula for counting the number of characters in a cell you can ask ChatGPT to “write a google sheets formula that will count the characters in a cell one column to the left” and it will create the formula. All you need to do is copy the formula and confirm that it’s been generated correctly.
Glossary
Just in case you feel late to the party and were apprehensive to ask about some of the basic acronyms, we’re including them here.
AI - Artificial Intelligence. At its most basic, AI is a concept where computers (artificial) act with human like capabilities (intelligence).
ChatGPT - Chat Generative Pre-trained Transformer. Chat as in it’s a chatbot - you can send it a message and it will send you a message back. Generative as in the tool is generating something, in this case text. Pre-trained as in the developers of the tool had to teach the AI how to behave and respond. Transformer is basically the framework that makes ChatGPT work, you can think of it as both the architecture and engine of the tool.
LLM - Large Language Model. When an AI tool has been trained using a huge amount of data. Tools like ChatGPT take an enormous amount of data that already exists, learn as much as they can about it, and use it to create content.
NLP - Natural Language Processing. This is a feature of AI technology that allows AI tools to understand the way that humans normally speak, and provide a response in a way that will be understood.