Learning & Teaching AI: A Chief Learning Officer's Journey
“Did you leverage your assistant?” This is a common question my learning team and I ask each other when discussing problems we are trying to solve, ideas we are trying to generate, or drafts we are trying to get out faster. That assistant does not get paid, and they don’t have to come to our weekly meetings because it’s not a human. It’s a large language model called ChatGPT.
It's been almost two years since we started actively exploring AI's potential. While some colleagues remain skeptical, others are enthusiastically sharing success stories. I fall somewhere in between, recognizing both the opportunities and the challenges.
AI offers the potential to automate tactical tasks, freeing up time for strategic work. However, it's equally crucial to understand the technology's nuances and address concerns about data privacy and potential inaccuracies. As a recent HBR article revealed, a startling 68% of participants chose not to edit the output of a language model, highlighting the need for critical thinking and responsible AI usage.
For me and my team, it is hard to imagine how we worked without this technology before. Generative AI has become a workplace staple, transforming how we work, learn, and operate. As learning and talent leaders, sometimes the biggest challenge is just knowing where and how to start. We started by asking ourselves what our role was in learning and teaching this technology. We came up with three primary responsibilities to ensure we are staying ahead and even leading the way with AI:
Upskill on AI Literacy: Understand what AI is, how it works, and the principles of prompt engineering.
Courageously Experiment: Actively use AI, share learnings, and demonstrate new ways of working.
Lead Through Change: Set a clear vision for AI adoption, educate on risks, establish governance, and recognize those using AI responsibly.
As part of a 3-part series, I’ll go into detail with each one of these areas. The first, and most important one is Upskilling AI Literacy.
Upskill on AI Literacy:
We can’t drive or lead change if we don’t know what we’re leading or changing. We agreed as a Learning and Organizational Development team that we would learn more about this technology together so that we could more effectively help other employees learn and use it as well. We leveraged our platform to take courses and share what we were learning each week in our team meetings. Our goal was to: “To build foundational knowledge and functional skills in AI and ChatGPT technology so that we can become responsible users”.
Over 3 months starting in Q1 of 2023, we talked through how we were using ChatGPT, exploring the technology in our work, and developing new team practices. Here are some examples of prompts we used in our first session together.
This one was based on work we were doing with assessments:
I manage the work of some contractors who review assessments my company is developing, and I want to send them a survey about their experience. The goals are (1) to identify any areas where I could improve the process or experience for them, and (2) to establish a baseline NPS metric for my job performance. Here's a draft of my survey. Any suggestions to help it better meet those goals?
Thank you! Can you suggest any ways to make the survey shorter?
We also made it personal.
Give me five healthy, easy, kid-friendly vegan dinner recipes using vegetables in season in California.
I don’t think she’d like those. Can you recommend some new ones that are more mild?
Great, thanks! Can you make me the grocery list for that? I want to make 4 servings of each.
We leveraged prompt templates like this one adapted from a course on our platform: ChatGPT 101: Supercharge Your Work & Life.
Use this template for your prompts: [verb(s)] [output and length] [task and objective]
An example could be:
Take on the persona of [expert persona]. The output should include [relevant data and resources]. The writing style is [tone of voice] targeted towards [audience]. Consider the following points:
[context]
Then we worked in groups of 3 to get ChatGPT to help us with something we have been working on, personally or professionally. We had to:
Pick one task for the whole group
Work together to experiment with prompts to try and get the best output (either screen share or work in parallel and discuss)
If you get a good result with one task, pick a new one and keep working
Then we had homework (or as we called it WorkWork) for us to do the next week until we met again. We listened to podcasts, shared favorite leaders in the field (Ethan Mollick became everyone’s thought leader to watch in this space), and also included challenges to like:
For one day (scheduled on your calendar!) Try to use ChatGPT for every single task.
Reflect: When did the human element need to come in? When was it helpful not to bring as much human element in to what you were doing?
We held each other accountable for this work by including it in our goals. See Lauren’s example below:
As we became more comfortable leveraging this technology, we wanted to upskill the rest of our organization. As we started to do more research on AI adoption, it was no surprise that a major hurdle to widespread ChatGPT adoption is the lack of adequate training. A survey in Denmark revealed that a significant number of workers, even those who recognize ChatGPT's potential to boost productivity, are hesitant to use it due to a lack of training. Anders Humlum, one of the paper’s authors and an assistant professor of economics at the University of Chicago Booth School of Business, shared that 43% of those surveyed cited the need for training as the primary reason for not utilizing ChatGPT.
A major hurdle to widespread ChatGPT adoption is the lack of adequate training.
Our CEO, Greg Brown, and the executive team made AI literacy a company-wide priority, requiring all employees to complete our foundational AI learning path by year-end. It included courses created by our very own Chuong (Tom) Do, Head of Engineering Data,
and one of our popular instructors, Axel Rittenhaus. While over 80% of our employees took the learning path, what was even more important were the conversations on teams, similar to those our team started having months before. We provided team discussion prompts as part of the learning path, encouraging people managers to:
Participate and complete this Learning Path
Ensure your team participates and completes this Learning path
Facilitate conversations with your team this month around Generative AI.
We broke down the learning path and experience by week and highlighted questions to discuss each week. See below for examples of that:
At our year-end company meeting, we showcased inspiring stories of how employees across various departments were utilizing AI to boost productivity and innovation. From streamlining tasks like summarizing employee feedback to accelerating code development, AI was making a tangible impact.
Today, we've taken this integration further with an internal GenAI assistant on Slack, powered by cutting-edge models like ChatGPT 4. This tool is designed to amplify creativity and efficiency throughout the organization.
As leaders, we recognize that embracing change starts with personal learning. It's our responsibility to continually explore new technologies and share those insights with our teams.
In my next article, I'll share specific examples of how we've experimented with AI, the unexpected benefits we've uncovered, and how we're leveraging AI to transform our learning experiences. Stay tuned for more insights into our AI journey!