To get the most from AI, you want to think about it in terms of the roles it serves in your daily workflow. The most important step of my Who-Why-What-How prompting framework is to assign the chatbot a role. And as I say often, the best way to harness AI is by creating custom GPTs, which are basically reusable, specialized roles.
So, let's look at the 10 roles AI should play in everyone's daily workflow. You can use them with any of the top general purpose chatbots, but I definitely recommend using that Who-Why-What-How framework. And they're especially useful when used in custom GPTs. That will be my next tutorial, so be sure you're subscribed to Productivity Nexus emails and on YouTube.
First up, we have The Copyeditor. Whether it's a newsletter, or a document for a client, or just an important email, I run every piece of text I write of any consequence through AI. I have a custom GPT preconfigured with the instructions:
You are an excellent copyeditor and an expert on English spelling and grammar. I will provide text, and you will identify spelling or grammar errors. You will NOT make stylistic recommendations; you will only identify spelling or grammar errors."
Then, I can tag the Copyeditor GPT and say:
Identify spelling or grammar errors in the following text."
I also ask it often about a single phrase. I'll say:
Tell me if the following sentence is grammatically correct. I'm mostly wondering about the word 'nurtured.' Suggest better alternatives.
I'm constantly learning from AI chatbots. Throughout each day, as questions emerge, I summon my team of AI tutors, which are custom GPTs that are primed for teaching and specialize in particularly subjects. For example, I have one for JavaScript and another for the Notion API.
So to configure them, I'll provide instructions like:
You're a professor of computer science specializing in JavaScript. I'm your student learning JavaScript. I've achieved intermediate proficiency. You will help me master JavaScript by answering my questions as I encounter challenges.
Summarization was my original favorite use of AI, and still use it to summarize content every day. You can prime any general-purpose chatbot for summarization, and then paste or upload the content. I like to request the key points as bullet points in compete sentences.
But my favorite approach is the Zapier Agents extension for Google Chrome, which can summarize the page I'm currently viewing, answer questions about the content, and send that information to other apps.
And for PDFs, Adobe's Acrobat extension can summarize, as well as analyze and derive insights from documents.
The Copywriter can compose content for a broad spectrum of needs and based on a variety of prompts and supplied resources. This is where the full Who-Why-What-How framework is important.
A good primer for this versatile copywriter is:
You're an excellent writer with perfect English grammar. You can write in any requested style, tone, or format. You excel at extracting key points from text for repurposing for other media.
Then you can submit a prompt like:
Below, I've pasted a guide to finding the top automation opportunities in your workflow. It's too technical and advanced for my target audience of everyday professionals with limited technological proficiency. You will simplify it by shortening it and using layman's terminology. In the style of Ernest Hemingway, you will use short, direct sentences.
I also use my copywriter GPT daily as a thesaurus. I'll say:
List 10 better terms for 'things' in the following sentence.
Along with ChatGPT, meeting assistants are one of the primary gateways into AI. They can transcribe and summarize and create action items from your virtual meetings in Zoom or Google Meet or other services. Afterward, you can chat with them to ask questions about and analyze your meetings.
I like Otter for its versatility beyond note-taking. Fireflies is another popular one. Granola has been praised recently for its approach to privacy.
You can provide AI various types of data for analysis, like sales data, website activity, or past performance of an investment your considering. With that data, AIs can:
They allow you to chat with data and request reports using natural language.
As a primer, you could instruct the AI:
You're a data scientist who excels at extracting trends and insights from any data set. From those trends and insights, you can make accurate predictions. I will provide data, which you will analyze to fulfill my requests and answer my questions.
And with tools like Zapier, you can connect AI to your data sources and configure reports on a regular schedule.
Here's one I wish my parents would use. I have a whole Geek Squad of custom GPTs for troubleshooting the tech problems we all encounter on a daily basis. One specializes in Apple products and another in the Google suite of enterprise apps.
Priming them is straightforward. I'll say:
You're an expert in Apple products, including macOS, iOS, and watchOS. You'll answer my questions to help me learn these operating systems and resolve issues that I encounter.
It's pretty amazing how they walk me through the resolution of most issues.
The Strategist is about as versatile as The Copywriter. They can reflect the roles of advisors in the real world, like advisors for your business or nonprofit; or financial advisors.
You could say:
You're a successful entrepreneur, and you coach aspiring entrepreneurs. As I build a business around a novel idea, I'll ask you questions with plenty of context so you can help make choices to maximize the success of my business. My key performance indicators are revenue and customer volume.
At a more granular level, AI can prioritize initiatives in your pipeline, compose strategic plans, and fully scope out projects.
The Researcher is the newest role in the mix but maybe the most impressive. And that's thanks to the Deep Research features that have rolled out recently by all the general purpose chatbots with access to the web.
What they do is, using that web access along advanced reasoning capabilities, they autonomously conduct comprehensive, multi-step research. They gather information from sources across the web, then analyze and synthesize it in a detailed report with citations.
You could say:
Compile a research report on the impact of AI on jobs, including past and present impact, as well as projected impact.
Lastly, regardless of your own role, you can benefit from AI image generators. At the very least, you can have a ton of fun with them.
With tools like DALL-E in ChatGPT, Google's ImageFX, and my personal favorite, Ideogram, you can create stunning photos, illustrations and even infographics that bring to life presentations, documents, websites, and any other material you produce.
In my video on the Who-Why-What-How framework, I used images generated by Ideogram to convey that context points the AI down the right path and constraints keep it on track.
Subscribe to Productivity Nexus emails and on YouTube for more insights around AI, automation, and systemization. 🦾