So you've decided to adopt AI within your organization. Now comes the hard part of rethinking your People, Processes and Technology. Getting the People component right is critical, and it's often overlooked in the rush to accelerate product development with AI.
As you start to train your engineering teams on AI-assisted and eventually AI-native software engineering, we have found it helpful to understand the archetypes that you will likely encounter in your organization.
There are 5 repeating archetypes that we see in engineering orgs:
- The Skeptic: typically a senior engineer who voices concerns about quality and maintenance of AI-generated code
- The Empowered: technical-adjacent people who have been unlocked by AI or senior engineers who are early in their AI-adoption journey
- The Cowboy: early career engineers who throw caution to the wind and want to YOLO AI
- The Blocker: a person who has taken a hard line against AI and has no interest in actively contributing
- The AI-native: a senior engineer who has put quality guardrails in place and is now able to 2x, 10x, and 100x their productivity and quality
The Skeptic
The Skeptic is usually a senior engineer who has genuine concerns about how to maintain quality and production readiness with AI-generated code. What is important about the Skeptic is that they are raising genuine concerns with good intentions. It's important to differentiate the Skeptic from somebody who is simply blocking progress (the Blocker).
Skeptics raise good questions that the organization should address. Questions around quality, code reviews, and architecture must be resolved to both increase velocity and maintain quality.
For example, The Skeptic will often ask about maintaining quality when the volume of AI-generated code is a mix of different quality levels, does not conform to architecture standards, and is being submitted at a volume that is overloading the senior staff.
As a leader, we've found it's helpful to engage with Skeptics, document their concerns, and then engage with them on putting a plan in place to rapidly address each concern.
The Empowered
The Empowered are technical-adjacent folks, such as SREs/Platform Engineers and Product Managers. Or they are experienced engineers who are simply early in their AI journey.
The technical-adjacent folks tend to be good technically, but have traditionally lacked the ability to write code beyond simple scripting. Or they are Product Managers who previously could not code a prototype. What is important about The Empowered is they are extremely diligent, and when paired with AI-native quality standards, can become hyper-productive.
For experienced engineers who have started to embrace AI, it's important to listen to the challenges they face. For example, most engineers who are mid-stream in their AI journey struggle to get consistent high-quality output from the AI. As a leader, help these engineers find solutions to each of the challenges they face to help them progress from Empowered to AI-native.
With AI, The Empowered can start to act independently. SREs can implement more advanced Infrastructure as Code (IaC) and PMs can generate prototypes. Experienced Software Engineers (SWEs) will start to see 2x and 5x productivity gains.
When working with technical-adjacent Empowered, provide them with guardrails when necessary, while acknowledging that some work is either throwaway (hello prototypes) or has a limited blast radius.
The Cowboy
Let's discuss the Cowboy. These are engineers who have not yet developed good engineering discipline and therefore want to YOLO AI. If left unmanaged, they will clog the PR queue with 10,000+ line PRs that don't follow any architectural patterns and will eventually slow down your product development.
The Cowboy, if left unchecked, will clog the PR queue which will overload the senior engineers.
The challenge with the Cowboy is you want to help them maintain enthusiasm while putting guardrails in place to ensure they don't nuke your codebase.
Managing the Cowboy looks a lot like managing any junior engineer. Help them develop fundamental engineering skills so that they begin to move from Cowboy to Empowered.
The Blocker
The Blocker is an engineer who has chosen anti-AI as a position and does not want to engage with AI initiatives.
It's best to keep Blockers on traditional projects rather than forcing AI upon them.
The AI-native
AI-natives are forged over time. While your team may not start with any AI-natives, they will emerge as Skeptics transition to Empowered and eventually to AI-native.
As a leader, focus on developing AI-natives and then help distribute their learnings and best practices throughout the Engineering org.
Wrapping up
Much as it's important to understand the AI maturity model, it is also important to understand the archetypes of people within your organization. Combined together, you can develop a plan that addresses people, processes, and technology so that your AI initiatives yield 2x, 10x, and 100x productivity improvements.
