Supporting your team on using generative AI tools

In today's fast-paced tech landscape, generative AI tools have become more commonplace within engineering orgs to boost productivity, creativity, and problem-solving capabilities. Love them or hate them, understand them or… don’t, these tools most definitely have the potential to transform the way we approach our workdays. However, like any powerful technology, generative AI comes with its own set of challenges and considerations.

The effectiveness of leveraging generative AI hinges on two critical factors: knowing what tools to use and understanding how to use them effectively. Without proper guidance and strategic implementation, these tools can become more of a hindrance than a help, and frankly, and be a danger to an org without proper controls. It's essential to strike a balance between leveraging AI for its strengths while maintaining the human element that is irreplaceable in creative and strategic decision-making.

Hence this newsletter! I’ve put a lot of thought into the use of generative AI and I also wrote our generative AI usage policy at Spot AI, so I wanted to share my thoughts with you here.

Generative AI doesn't replace junior engineers. Charity Majors recently wrote about this for Stack Overflow, highlighting the importance of junior engineers. I’m not going to repeat it here; just read it. Generative AI should complement, not replace, the need for developing talent. Junior engineers bring fresh perspectives and are crucial for long-term team growth.

Understand organizational policies. Before leveraging generative AI, ensure you have permission. Familiarize yourself with your organization's generative AI use policy, if they have one. (If they don’t, ask them to write one!) This is crucial to stay within acceptable use boundaries and to avoid potential legal and ethical issues.

Use generated code as a guide, not gospel. Generated code is a great starting point for guidance or handling mundane tasks but rarely is production-ready immediately. I often use generative AI for fun coding tasks in my side projects. It typically takes a few iterations to get it right, and sometimes I end up spending time fixing what was generated. Always review and test the code thoroughly before deploying it to production.

Protect PII and IP. Avoid entering personally identifiable information (PII) or intellectual property (IP) into generative AI tools. For instance, if working with JSON objects, obfuscate any customer data. Research how the AI tool's team handles customer inputs for retraining to ensure data safety. If they re-train based on customer inputs, it’s super important to obfuscate key data.

Encourage experimentation, but set boundaries. Encourage your team to experiment with gen AI tools to discover new efficiencies and solutions. However, set clear boundaries to ensure that experimentation doesn’t lead to dependency. Avoid having them create bad habits. Balance is key; use AI to enhance human creativity and problem-solving, not replace it.

Invest time in research. Make sure your team has access to training and resources on how to effectively use gen AI tools and what to look for in a good or bad tool. This includes understanding the strengths and limitations of these tools, as well as best practices for integrating them into their workflows.

Promote ethical use. Create a culture of ethical use of generative AI within your team. Discuss the potential biases and ethical implications of AI-generated content. Ensure that the use of AI aligns with your organization’s values and ethical standards.

Generative AI can be useful, but only when used safely and ethically. Lead the way on advocating for a culture of positive behaviors when leveraging gen AI. Remember: Engineering isn’t the only department that wants to leverage generative AI, but engineering can set the best example of how the entire company should operate.