Using generative artificial intelligence (AI), including large language models in the workplace can influence how teams operate, offering a range of benefits from streamlining processes to driving innovation. By incorporating AI tools into daily workflows, teams can work together more efficiently and effectively. 

We implemented our own AI Education Pilot Program at Atlassian, spearheaded by Alfredo Huitron, a Product Manager at Atlassian, which incorporated a mix of tools, including generative AI. During this program, we had the opportunity to observe 25 project managers and knowledge workers at Atlassian over 6 weeks as they tested and implemented AI into their daily routines. Through this experience, we learned valuable lessons about educating teams on the effective use of AI in their everyday tasks. Here are the top 10 lessons we learned: 

1. Learn AI’s strengths and weaknesses

Understanding what AI is good at and what it struggles with is essential for effectively integrating it into daily workflows. Recognizing AI’s strengths — efficiency, speed, and the ability to manage large volumes of information — allows teams to assign specific tasks they know align with them. For instance, AI’s ability to swiftly process data makes it well-suited for tasks requiring quick analysis or repetitive operations.

Acknowledging AI’s weaknesses, such as limitations in deep thinking and reliance on context, helps teams set realistic expectations and make better decisions about when and how to use AI solutions. For example, teams may find AI unsuitable for creative problem-solving tasks, such as generating original content for marketing campaigns, designing innovative product features, or conceptualizing new business strategies. 

In these scenarios, human creativity, intuition, and domain expertise are essential for generating unique ideas and solutions that resonate with customers and drive business growth. AI isn’t as creative as human professionals, making it less effective for tasks requiring originality and out-of-the-box thinking.

2. Assign AI specific tasks

AI tasks must align with its capabilities. Teams can streamline processes and free up valuable time for more strategic activities, but AI prompts must be specific enough to provide AI solutions with the context they need to perform those tasks effectively and accurately. Whether generating reports, reviewing analytics and data, or drafting emails, assigning AI tools specific tasks ensures they’re used efficiently within the organization. 

Additionally, providing clear instructions and context for each task ensures that AI delivers accurate and relevant results. The more specific the task, the better AI can understand and execute it.

3. Iterate quickly

When working with AI, start with small tasks. This way, you can see what the results look like and adjust as needed. It’s okay if things aren’t perfect at first — AI learns from each interaction. Over time, you’ll get better at asking for what you need, making the process smoother and more natural. With practice, using AI will feel as easy as searching the web.

4. Keep in mind ROI

When using AI, ask yourself: Is it making your work faster or better? If not, it might be better to handle the task yourself.

Consider how quickly you’re getting valuable results. Are you spending too much time refining prompts or editing outputs? If using AI takes longer than doing it manually, you need to reassess. 

5. Learn basic prompt engineering techniques

Your AI prompt ideas should be clear and provide context while focusing on specific tasks. For instance, if you’re using AI solutions as part of your marketing strategy, try using them for small tasks like drafting emails instead of tackling the entire campaign in one conversation. 

Large, broad tasks, like asking AI to develop an entire content strategy, often result in generic outputs. However, by narrowing the scope of your request, you can get a more helpful response faster and with less context required. 

6. Add context and instructions to your prompts

Again, the best AI prompts give the AI tool context, but you should also provide it with instructions. When drafting emails or generating content, include relevant details such as the target audience, key messaging points, and desired tone. The more information you give AI, the better the results will be. 

By offering precise instructions on the content structure and formatting, you can ensure that the AI solution generates outputs that align with your goals. 

7. Select the right AI tool for the job

You can use various AI capabilities in tools like Confluence within your organization but remember that each is designed to address specific tasks and challenges. When selecting an AI solution for a particular task, consider factors like functionality, scalability, compatibility with existing systems, and how easily it can integrate into existing processes. 

Whether you’re analyzing data or generating content, there’s likely an AI tool available to meet your needs. By finding the right tool for the job, you can use AI to streamline workflows and improve innovation and collaboration

8. Experiment to get the best AI outputs

Experimentation allows you to try different tools to determine the best option. You can get the best outputs by experimenting with various techniques, approaches, and prompts. For example, provide more context or clearer instructions to see if it improves the quality of the AI-generated content. 

You should also iterate on your experiments. Don’t be afraid to make adjustments and try again if you don’t get the results you’re looking for initially. 

9. Understand how and when to use AI

AI is not a one-size-fits-all solution; misusing it can lead to subpar results or negative outcomes.

One key consideration is task complexity. Some tasks are well-suited to AI automation, while others require human judgment and intervention.  

Meanwhile, tasks that involve complex thinking and nuanced decision-making, such as strategic planning or creative problem-solving, may be less suitable for full automation with AI. While AI can assist with these tasks by providing insights or suggestions, the final decisions often require human input and expertise.

Another factor to consider is the availability of data. AI requires data to learn and make decisions, so tasks that involve large amounts of structured data are often well-suited to AI automation. 

10. Manage task complexity

AI is most effective when applied to tasks that are well-defined and of manageable scope. Trying to tackle overly complex tasks with AI can be challenging, leading to poor results. 

One approach to managing task complexity is to break large tasks down into smaller, more manageable components. By dividing complex projects into smaller, more digestible tasks, you can make it easier to apply AI effectively. For example, instead of automating an entire business process with AI, focus on automating individual steps or sub-processes.

Another option is to prioritize tasks based on their suitability for AI automation. Tasks that are repetitive, rules-based, and involve large amounts of data are often good candidates for automation with AI. Conversely, tasks that require creativity, judgment, or human interaction may be less suitable for AI automation.

Additionally, it’s important to consider the level of expertise and resources required to implement AI solutions effectively. Some AI techniques and technologies may be more accessible and easier to implement than others, so it’s essential to choose the right approach based on your organization’s capabilities and goals.

Adopt artificial intelligence into your everyday operations

Using artificial intelligence streamlines processes and boosts productivity. With AI solutions, you can improve decision-making, save teams time, and drive innovation across the organization. By incorporating AI tools into your daily workflows, your team works smarter, not harder.

Top 10 things we learned about educating our teams on using AI every day