Five Ways to Build a Culture of AI Experimentation

As leaders, we're trained to have the answers. We set the strategy, the roadmap, the headcount, the resourcing. Answers are a fundamental part of the job. Or at least that’s the common misperception in leading teams. We think we need to have answers. This is especially true with artificial intelligence (AI), when the buzz is palpable and our teams are clamoring to understand “what does this mean for me?”

So it’s no surprise to us that while a recent Slack Workforce Index shows that the share of leaders looking to incorpora te AI tools into their business has grown 7x since Sept. 2023, over two-thirds of workers have never used AI at work.

We are all searching for answers: What tool should I use? How should I use it in my job? Is it safe and secure? What other considerations or guardrails should I place? 

The two of us have led teams across research, analytics, product, and strategy whose jobs are to chart a path through ambiguity. But if there’s one thing that we’ve learned in our careers, it’s this: when we are so early in this journey of figuring out AI at work, the best path forward is to create a culture of experimentation rather than have specific answers.

We have to empower our teams to think like researchers. To realize the full potential of AI, companies need to create a safe space to experiment with AI and empower workers to participate in shaping how AI can be applied to their specific jobs.

Here are five ways leaders can start to create a culture of experimentation with AI: 

1. Put in the work to make AI work. 

Only 15% of workers globally strongly agree they have the education and training necessary to use AI effectively. It’s important to provide guidance on how to get started and to carve out time and space to experiment and evolve how AI might be used in their daily tasks.

Across both Slack and Women Defining AI, we have run a lightweight experiment where we provide 10 to 15 minutes of structured education daily for three weeks. Participants are then given a short, hands-on project where they can apply what they’ve learned. More than anything, it allows people to try using AI in an approachable way and build intuition in different applications, whether it's in their personal life or at work.

Expert tip: Incorporate quick sprints of weekly training sessions (they can be as little as 10 minutes) that allow employees to get their hands on AI and get used to using it. 

2. Build comfort with “this didn’t work.”

Fostering a culture of AI experimentation means creating an environment where employees can explore, innovate, and learn without fear of failure or retribution. Workers who feel trusted by their employers are 94% more likely to have tried AI for work-related tasks. Key to integrating AI in everyday work is giving employees the freedom to take risks, providing them with the resources they need like AI usage guidelines, and making sure they have the time to really be creative in their ideas.

In a recent guide to AI in the workplace, Charter described how Coursera CEO Jeff Maggioncalda embraces vulnerability and shares the AI prompts he is testing with employees. The prompts don’t always work as expected, but his approach inspires employee trial-and-error. 

Our favorite Women Defining AI examples are when members take projects to unexpected places, like one who used AI to not only create a new Dungeons and Dragons (D&D) character but also an entire play script to use in her D&D meetups. Another member uploaded custom information about her style guidelines and preferences to create a custom AI bot that suggests and assesses work outfits for her day.

It’s through these more creative use cases that our teams create novel approaches where AI can be useful, demonstrating how much brainstorming or ideating AI can really tackle (as in the D&D example) or how much you can personalize an AI’s responses with a simple file or two (as in the work outfits example).

Expert tip: Host frequent workshops or forums to create spaces for your team to exchange ideas and test AI experiments with colleagues. Share both the ideas that worked and those that didn’t work and discuss why.

3. Embrace incremental change and start small.

If AI is going to completely transform work, it isn't going to happen overnight. Such change also feels out of reach for many employees. Instead, start small. Chip away. AI can transform your job, little by little.

At Women Defining AI, we often see that the easiest way to start is to identify “one thing you do frequently at work or in your life” and then experiment with different AI tools to get that task done.

Expert tip: Be specific and start with experiments with simple use cases within a small team. When you get specific, it's easier to know what to experiment with and you can assess whether AI is actually helpful with this task. The goal is to then repeat this process over several core tasks in your workflows. Over time, you will build an intuition about the ‘aha use cases’ where AI can help make each task more efficient and, more important, unlock completely new ways of approaching them.

4. Cultivate collaboration and knowledge sharing.

We are all collectively still finding novel ways to use AI, as teams continue to experiment, and the underlying technology continues to improve rapidly. Sessions or online chat spaces where people can share their AI projects and findings to the wider organization disseminate valuable insights and experiences. This knowledge exchange stimulates creativity, inspires more ideas, and encourages continuous learning across the organization. 

The Women Defining AI member who made that AI style bot inspired others to pursue ideas, including an AI bot that can provide a concise explanation of trends in data and charts and a bot that gives feedback on the layout of web page designs. 

Slack Workforce Lab created an #ai-lab Slack channel where any employee is encouraged to share their experiences of integrating AI into their daily tasks. To incentivize sharing, the team hosts monthly AI challenges where the most innovative or inspiring use case or learning wins a prize. One recent challenge winner used AI to provide an executive summary of a work-related podcast episode, saving one hour of time to consume a lot of information. 

Expert tip: Create an open Slack or Teams channel or other communication workspace that encourages teams to share how they are using AI and spotlights real-life applications of AI in the organization. Encourage front-line managers to discuss ideas within their teams and spotlight great ideas proactively in different forums like a team channel or an all-hands meeting.

5. Reimagine how you measure impact.

What gets measured, gets managed. As we build a culture of experimentation and encourage people to find ways to improve work with AI, we also have to change how we measure impact. 

The Slack Workforce Index found that employees are more likely to spend their time saved from AI on routine administrative tasks rather than high-value activities, like skill-building and creative projects that could lead to more innovation and improved customer service for the business. That’s a red flag for leaders and one they have the power to address. 

If leaders continue to measure productivity through the lens of activity or input metrics—such as time spent on a task—employees fill any time saved with performative work that doesn’t contribute to company and team goals but is done to appear productive. Instead, focusing on defining clear outcomes—such as customer satisfaction or developing new product offerings—will encourage employees to shift from mere busywork to using their time to be creative and employing whatever tools they have to achieve those outcomes.

Leaders can also create incentives that support experimentation and innovative ways to improve work. For example, Skillsoft, an educational tech company, holds organization-wide innovation challenges and rewards innovation with spot bonuses and recognition in performance reviews. 

Expert tip: Move from tracking activity and input metrics like time in office and number of tasks completed to incentivizing outcomes like customer satisfaction and time to market. Encourage colleagues to use AI tools to pursue those outcomes, even if it looks like they’re doing less “work.”

In this AI revolution, it's not just about the tools. Making the most out of AI starts with workers. When people are empowered to experiment with AI, they’re more likely to adopt a mindset of continuous learning, creativity, and problem-solving. This culture of innovation then becomes the driving force behind new solutions, reimagined customer experiences, and business growth.

Written by Christina Janzer and Helen Kupp.

© 2024. This article is reprinted with permission by Charter Works Inc. All rights reserved.