You’re reading this because you know artificial intelligence (AI) has value, but you don’t know how to use it. This blog gives you the 7 ways to use AI in your work. These use cases can be used with any AI model including ChatGPT, Gemini, Claude, or Grok.
All of the use cases below assume you are familiar with how to use AI tools. If you want a basic overview of how AI works and how you can use it, check out our article The Basics of Using AI.
Table of Contents
Content creation
The highest return on investment for AI is content creation. Content creation doesn’t mean you are a “creative” such as a graphic designer, podcaster, or videographer. It means you create content of any kind. If you’ve written an email, you have created content. That’s almost everyone.
Content creation is easier with AI for one simple reason: it is more efficient to start with a rough draft than a blank page. That’s true for anything. You could be creating an email, year-end report, text message, resume, product description, policy document, speech, sales talking points, or countless of other deliverables and it will always be easier when you start with something rather than nothing.
Whenever you prompt AI, we recommend using the CASE format:
- Context: Start your prompt with the story of your ask. When someone asks you to do something, there’s usually a story that comes with it. What are you working on? What’s your role? Why are you doing this? Is there a deadline? Not all context is necessary, but some can provide direction.
- Ask: State what you are asking AI to do. This is the most powerful part of your prompt. No matter the length of your prompt, AI is looking for your ask. If you write a book-length prompt then end with the ask “respond with a short compliment,” then you will get a short compliment. When it comes to AI, ask and you will receive.
- Samples: Share a sample of what you want — such as an attached document. Samples give AI direction without you taking the time to explain it. A sample document can give direction on tone, formatting, style, and many other creative decisions. You don’t need a sample for every prompt, but they can be very powerful for your first prompt with an AI.
- Examples: Tell the AI how it can succeed. This is not the same thing as Samples, but the two work together. A sample is a document. An example is your expectations. Examples are prompts like “make it sound professional” or “format this in a list of bullets.”
Your context, ask, samples, and examples don’t have to be a single sentence. You can prompt AI the way you would chat with another person — casually, with loose direction — so long as your overall message is clear. Here are some examples:
- I just started a new job and I want to introduce myself to my manager. I’m worried I’ll make a bad first impression. Write an email introducing myself and convey how I can provide value to the team. I’ve pasted below a previous email I wrote to give you an idea of my writing style. Make me sound professional but approachable.
- I am writing my goals for this upcoming year to share with my team. Last year we had layoffs and lost 30 percent of team of 26 people. I know there’s a big shift toward adopting new technology like artificial intelligence. Me and my team work in sales and some of my colleagues are thinking about how to implement AI into our process. Last year I was evaluated primarily on the amount of outreach I did. I contacted more than 5,000 leads. Considering our team reduction, I think this year’s expectations will be larger. These seem the most relevant concerns, but I don’t want to get tunnel vision on my concerns. I want to include some goals that are common across industries. Draft a goal statement in the format of bullets. I’ve attached my own personal goal statement from last year. For each goal with a short description of the goal. Use the best practices and standards in the healthcare industry for this type of deliverable.
- A professional acquaintance keeps reaching out to catch up. They’re a nice person, but I really don’t have time to meet with them. I want to say this to them without hurting their feelings. Write a text message response that is polite but indicates how I feel. I don’t want to get invited again.
Remember the output you get from AI is often a good first draft, but rarely a great final draft. The current state of AI relies on you to provide your skills and experience to ensure AI is aligned with your goals.
Summarization
The most common usage of AI is summarization. This is for one simple reason: everyone wants to read more, but they don’t have the time.
People don’t read because there’s too much to read. Research shows the average office worker spends 600 hours a year reading emails. That’s almost a third of your work hours in a year (2080 hours a year). This data doesn’t include reading other documents. You probably read other things like project updates, draft deliverables, memos, chat messages, text messages, etc. It’s not uncommon for “work” to be synonymous with “reading email.” With all those hours going to reading things, when do you have time to do your job? For many people, that is their job. It’s not rewarding or efficient.
Take back your life by getting summaries from AI. Remember, the format of “context, ask, samples, and examples” are just as important when requesting summaries. If you’re attaching a 50-page document, the AI doesn’t know how to prioritize the information unless you tell it your priorities. Here are some examples:
- I’m reviewing a deliverable by my team. This is supposed to be a short blog entry that connects with the reader with direct language, practical examples, and plentiful screenshots. Provide a summary of the information in this deliverable and inform me if these requirements were met.
- Look at this email chain. Give me a brief summary of events but include if 1) a request was made of me 2) a request was made of my manager 3) there is some sort of block to completing the project being discussed. I do not need details if this is a low-value back and forth with little information exchanged.
- I have been on vacation for 10 days. Review my inbox and provide a summary of the most important emails I received in that time. Important can be defined by: an email from my manager, a request made of me, a time-sensitive deadline of any kind, or any other definition of “important” you are familiar with. You do not need to inform me of other emails outside this criteria.
Two of the above examples are specifically for summarizing email, which is a unique strength for AI tools that integrate with email clients such as Microsoft Copilot or Anthropic’s Claude. Check out our article comparing the differences of each AI model.
If you don’t have an AI that integrates with your email you can still take advantage of these prompts by attaching or copy/pasting emails in your prompt.
Extraction
One of the quickest improvements to quality of life is relying on a technique called extraction. Extraction takes information from somewhere and places it somewhere else in a different format. It sounds simple, because it is! Extraction is most common when you’re taking information/data from multiple sources and compiling it for a new document.
For example, if you are assembling a report that relies on information from different places. This could be yearend financials for a company, analysis of housing costs in a market, or collecting data on consumer trends. The information driving these reports may come from 10 different sources, using 10 different formats, and may even use 10 slightly different metrics. One might report overall costs, while another reports costs per capita. One report might be a PDF and another is a spreadsheet. Collecting this information into one document can be incredibly time-consuming. Extractions saves countless hours of copy/pasting and reformatting.
- I am assembling a year-end financial report for my company. I’ve attached 4 documents detailing financials for 4 different departments: Sales, Operations, Customer Support, and Marketing. Extract the total costs from each document and create a table identifying the department, employee headcount, total costs, and total costs per employee.
Classification
One of the greatest values of AI is getting neutral and unbiased analysis. You can use AI to classify the sentiment or tone of a document. This is especially valuable when you’re working with large datasets or when you’re looking for feedback on your own creative content.
Humans are naturally prone to a variety of biases and that can confuse your understanding of data. For example, if you’re looking at customer feedback on a new product launch you may see the first 10 reviews out of 100 are all explicitly negative. We know from established science that reading 10 reviews back-to-back is going to influence your interpretation of the data. You’re going to be biased to think the overall feedback negative, but that may not be true. The next 90 reviews could be positive, but those first ten are going to stay with you. Asking AI to analyze datasets allows you to overcome this bias and work for productively.
Classification is best used for datasets like the above example, but it can also be used for general feedback. An emerging use case for AI is mental health support to reveal insights about your own behavior. You can keep a personal journal for your own mental health, then provide this journal to AI and ask for feedback on reoccurring issues or challenges you need to address in your life. Of course, beware of the implications of providing AI personal details about your life.
Examples:
- I am reviewing consumer feedback on our new website. Review the attached document and classify the general sentiment of the feedback. Organize this output into positive, neutral, and negative. For each sentiment, include the top three topics/concerns brought up by these sentiments. Provide this in a bulleted list format.
- I’ve written a blog entry explaining 7 ways to use AI in everyday work. Review this blog and classify the tone and style of the writing. Provide your analysis of who this type of writing appeals to the most. Additionally, provide feedback on how the tone and style could change to fit the audience of small to medium sized business owners.
Translation
Translation is an underutilized feature of AI. Most people rely on Google Translate for their minimal translation needs, but existing translation services have their challenges. Any bilingual person can tell you how literal translations can fail to capture the true meaning behind slang or colloquial phrases.
Imagine you’re reading a letter from an international business associate who is happy to report their department “keeps money in the dark” because a literal translation of financials being “in the black.” Or when a word is used in one language that has no true translation in another language – words like [continue later]
The technology driving AI is called a large language model. We have a full technical breakdown of what that means, but the important thing to understand is it is based on language. This uniquely positions AI to translate to any language on the planet – well, most of them.
Use translation to expand the reach of your content.
- I’ve created a blog explaining the value of artificial intelligence, but I want to translate it to other audiences interested in this topic. Review the document I’ve attached and translate it into Spanish. Identify any sentences or sections that needed to be significantly re-written to retain their meaning after translation.
Editing
When it comes to editing with AI, remember to dream a little bigger. You don’t need an AI model to implement edits like grammar or spellcheck. That form of editing was already available in inexpensive products that don’t use AI technology. Editing with AI enables you to implement large sweeping creative changes. With a simple prompt, you can rewrite a deliverable to fit for a different audience.
Examples
- I’m rewriting product descriptions for my company’s offerings. These products were originally for young professionals and have a casual, hip, and affordable value. We’re pivoting to appeal to older professionals and we want to communicate seriousness, reliability, and high quality. Rewrite the descriptions of these products for this new audience. Keep the length of descriptions similar to the previous ones.
- My manager gave me feedback on this deliverable. Implement her edits. I’ve pasted her notes below.
“Too wordy. Too many technical details. Use more storytelling at the beginning. Include a call to action between every third paragraph. This is decent but needs work!”
Problem Solving
The final use for AI is the most broad and has the highest potential value. Since this technology effectively unlocks all of the world’s knowledge, this means you have the opportunity to solve any problem in your work even work beyond your personal skillset and experience.
In pre-AI work, most workers specialized in a skill and relied on coordinating with others to effectively tackle large-scale problems. With AI, you can become a specialist in any discipline.