Less talking about AI, more using it

(This post was written by a human with a bit of help from AI.)

How can you do real things with AI at work?  

If you’re stuck in the ChatGPT dabbling phase (i.e., writing emails etc.) and want to sink your chops into something meatier, read on.  

Below, we’ve summarized the mind-blowing takeaways from our recent webinar, Learn to do more with AI: Practical tips beyond ChatGPT.  

In one hour, our two amazing guest speakers, Helen Lee Kupp and Nichole Sterling, the co-founders of Women Defining AI, walked us through SIX real tasks that all business and HR leaders could face in their day-to-day work, showing us step-by-step how AI could speed up each one.  

And while ChatGPT is included among their examples, the speakers opened our eyes to new possibilities with this tool.  

‍

AI Optimism

The first step is being open to using AI. Sure, there are lots of unknowns, but it’s here to stay, so your best bet is to jump aboard. That’s why we were very pleased to see so much optimism about AI among our webinar attendees.  

Just check out this word cloud from the event—out of 249 responses, the most popular words from attendees used to describe how they feel about AI were “excited” and “curious” (albeit with a sprinkling of “uncertain” and “scared”).  

‍

‍

It’s a good sign because it’s much easier to use a tool if you’re EXCITED and CURIOUS about it rather than uncertain or scared, right?

‍

So, Who’s Using AI? (And Who’s Not?)

In case you think AI isn’t for you, think again. Everyone will need to use it—women and men in all roles at all seniority levels. If not, you could fall behind as the technology continues to evolve, changing and creating jobs in its wake.  

And guess who’s at greater risk of falling behind? Women.

• One Harvard study found that on average, women use AI 20% less than men.  

• And, according to Charter, a study of all generative AI courses offered by Coursera found that globally, men comprise 68% of the course enrollments compared to 32% of women.

Why the disparity?‍

Helen and Nichole, whose mission at Women Defining AI is to close this gender-based technology gap, attribute it to a number of things—everything from women thinking they're not technical enough to feeling they're “cheating” if they use AI.  Another factor, women tend to follow the rules more and want detailed training before they try something, whereas men tend to be more open to experimentation. The Economist [note: subscription required] found that even high-performing women are hesitant to use AI at work because they feel they have to grit their way through it and do the hard work without help.  

‍

AI’s Impact on Entry-Level Jobs

It’s not only women who are at risk of being left behind in the AI revolution. Those early in their careers could be impacted, too.  Helen and Nichole highlighted a sobering US study of active job postings (Canada mimics the US) that shows a huge drop in available entry-level jobs. From January 2022 to July 2024, entry-level job availability dropped from 100 to 64.5, whereas other jobs held at 101.

“This has been affecting folks from customer service to software developers, where more enterprises and workplaces are starting to automate some of those workflows and those jobs,” says Nichole. How are these young workers going to get the required skills to advance in their careers if they’re not given that first job?  

‍

Practical AI Use Cases You Can Implement Today

Helen and Nicole emphasized that AI is more than just a search engine—it’s a tool that can help solve real-world business problems. Here are six use cases they showcased:

‍

1. Role-Playing Difficult Conversations with AI

We all have to have difficult conversations at work—delivering performance feedback, terminating an employee, negotiating for raises or job promotions, and more. Nichole showed us how to use ChatGPT’s advanced voice mode (paid subscription model) to rehearse these conversations ahead of time AND receive feedback, so you can improve your delivery.  

“Difficult conversations are part of the job,” says Nichole. “But practising with AI before having them with real employees? That’s a game changer.”

Use case: Nichole role-played a tough HR conversation with “Alex,” a difficult employee (AI-generated) who is resistant to feedback.

Tool: ChatGPT’s advanced voice mode (available with paid subscriptions)

Prompt they used:

“Hi Alex, we’re going to do a difficult conversation role play, where I'm going to be the people and culture or the HR professional and you're going to be the employee, and I want you to be really resistant. I want you to come up with some excuses about your behaviour, make it challenging for me, and let's see if I can help you see the impact of your behaviour by going through the difficult conversation process. Sound good?”

Tips:

1. At the end of the conversation, ask ChatGPT to provide two areas where you could improve on the conversation.

2. Use this tool to practise cold calls, benefits discussions or customer negotiations. You can even use it to think out loud, speaking into Advanced Voice and having it synthesize or summarize your thinking.

‍

2. Building Custom AI-Powered Tools—No Coding Required

Tools like Claude.ai can help even non-techies create basic, interactive tools with no coding skills required.  

Helen showed us how Claude.ai’s Artifacts feature (available for free) can generate a fully functional total compensation calculator in under a minute with a simple prompt.

“You don’t need to wait for engineering support to build these tools,” Helen explained. “AI makes it accessible for everyone.”

Tool: Claude.ai (Artifacts)

Prompt they used:

“Please create a Benefits Customization Tool that allows HR to: Adjust an employee’s base salary and see how it affects overall compensation, Toggle health insurance coverage and provide three different plans, Adjust RRSP contribution percentage, Customize paid time off days, Opt for additional benefits like gym membership and add any benefits you can think of that can be toggled on and off.”

Watch Claude.ai build the tool in the right pane.  

Tips:  

• Click the publish button at the bottom of the artifact to share the tool with your team (it generates a link).  

• If someone shares an artifact WITH you, you can customize it to your liking by adding features or updates to their original idea.

‍

3. Policy Creation and AI-Assisted Editing

We hear you, writing policies is soooooo time-consuming. Here’s how AI can help.

Nichole showed us how ChatGPT’s Canvas feature (or any of a number of free AI tools like Claude.ai, Gemini or Copilot) allows users to quickly create a first draft of a new policy, and then refine it by adding at the bottom of your prompt additional context or specific laws/guidelines that you want the AI to incorporate.

Use case: Draft a parental leave policy, making live edits and updates based on AI recommendations.

Tool: ChatGPT’s Canvas feature (or any of a number of free AI tools like Claude.ai, Gemini or Copilot)

Prompt they used: “We need to write a parental leave policy. Please include things like time off during the birth or adoption of a child. Our policy will entitle leave for 20 weeks. Also, propose other important aspects of the parental leave policy.”

Tips:

• Try two different tools to get a sense of variations in style and output.

• The more specific you are with your prompt, the better the first draft will be. For example, add context or background materials on the jurisdiction or laws that might govern your policy.

• Upload prior examples of policies you like for it to build on.  

‍

4. Perform an Impact Assessment

When drafting a policy (e.g., parental leave), how do you know it’s going to reflect all employees in your organization, taking into account their particular life or career stage and personal situation? Sure, you can ask multiple people (best practice), but this can be time-consuming and potentially sensitive depending on the topic.

Nichole and Helen showed us how you can use ChatGPT’s Canvas feature to test different perspectives when developing a policy and identify any blind spots.  

Use case: Provide different perspectives on my parental leave policy, highlighting how it would impact employees of all levels.

Tool: ChatGPT’s Canvas feature (or any of a number of free AI tools like Claude.ai, Gemini or Copilot)

Prompt they used: “Please identify the impact of a parental leave policy on all levels of employees, from entry-level to executive, and across different departments. Are there perspectives we’re missing?”

Tips:

• Upload the policy you want to get perspectives on.

• Add to your prompts the specific employee groups or scenarios you want to consider. For example, single or same-sex parents, adoption, fertility treatments, etc.

‍

5. AI for Research, Meeting Prep (and a Surprise)

So much reading to do, so little time, right? What if someone could summarize dense, long reports for you and deliver the top highlights in an engaging, podcast-like format that you could listen to on the go?  

Hello, Google Notebook LM, lovely to meet you.

“I use this tool to digest long reports while I’m commuting,” Nichole shared. “It turns dry text into something engaging and easy to absorb.”

It’s very handy to prep for meetings, too, helping you shine by quickly summarizing the key points you need to know. Best of all, Notebook LM is free with any Google account and does not train its models using your data by default.

Use case: Summarize a long article into an engaging podcast.  

Tool: Google Notebook LM

TRY IT: In Notebook, click Create a New Notebook, and then upload a source that you’d like to summarize (i.e., files, links, or pasted text). Watch it turn your file into a podcast (using two default voices; you can’t customize them).

FUN SURPRISE – YOUR CAREER AS A PODCAST: Imagine listening to a podcast about YOUR OWN CAREER (or someone else’s)!  

1. In LinkedIn, on your profile page, click Resources > Save to PDF.  

2. Upload your PDF to Notebook LLM and presto!  

3. In minutes, it will create a podcast version of it that you can download and share (as a .WAV file by default).

‍

6. Creating a Personal AI Assistant for HR

Yes, this one’s a bit more technical, but you don’t need any coding experience.

One of the most exciting demos was how to build an automated HR handbook (think, HR in your back pocket) using ChatGPT’s custom GPT platform (paid subscription required).

“This isn’t replacing HR,” Helen clarified. “But it’s an effective way to reduce repetitive inquiries and provide immediate, accurate answers.”

Use case: Create an HR handbook that new hires and anyone in the company can consult to answer common questions about time off, vacation allowances etc.  

Tool: ChatGPT’s custom GPT platform (paid subscription required)

Steps: This one isn’t a straight up prompt, so we’ll explain how it’s done.  

1. In ChatGPT’s custom GPT platform, click Create GPT, and then click Configure.  

2. Type the following info:  

• Name: “Company Handbook”

• Description: “A handbook that details benefits, leave, etc.”  

• Instructions: “This GPT will answer questions from employees on benefits, procedures, policies, etc.”  

3. Upload any policies/documents you’d like it to reference. For example, Helen suggests uploading an employee handbook, Mission statement, Vision, Values etc.  

4. Pro tip: Always add the text, “Please confirm answers with your HR professional” so your employees know not to rely wholly on this tool and to always check answers with HR.  

‍

Prompt Engineering 101

As you can see through these examples, the more specific you can get with prompts, the better the output. And it’s not one and done—it’s best to keep iterating and asking follow-up questions, so the AI improves and refines its output.  

“Treat it like an intern is my biggest advice,” says Nichole. “Think about when you give an assignment to an intern, you want to be pretty clear, pretty explicit. It’s the same thing with AI.”

Tip: You can even ask it to offer feedback on your prompts, so you can improve too.

Here’s a cheat sheet from Helen and Nichole:

What’s Next? AI Agents and the Future of Work

Helen and Nichole also highlighted what’s on the AI horizon—AI agents. Unlike chatbots or assistants, AI agents are designed to autonomously perform tasks like scheduling, traffic management, and workflow automation.

“AI agents will redefine how we interact with technology,” Nichole predicted. “They’ll move beyond simple task completion and start making real decisions.”

‍

What’s Next for YOU?  

1. Start experimenting. AI is best learned by doing—whether it’s practising a conversation, building a calculator, or using AI for meeting prep.

2. Stay secure. When possible, use paid versions of AI tools because you can control data retention and training. Regardless of the tool, always avoid inputting confidential information.

3. Think beyond chat. AI isn’t just for answering questions—it’s a tool for automation, simulation, and business strategy. AI is your assistant—designed to help you do more in less time.  

‍

As AI continues to evolve, staying informed and hands-on will be the key to unlocking its full potential. (Yes, that line was written by ChatGPT. Can you tell?) Questions or comments? We’d love to hear from you: hrtopics@altis.com.