Engineering Leadership
Leaders can shape culture through two levers: policy and daily choices
Systems thinking reveals how organizations adapt and interconnect—and why AI rollouts often amplify problems elsewhere in the system.
InfoQ
In a recent conversation, Elisabeth Hendrickson and Joel Tosi explored how viewing organizations as dynamic, interconnected systems offers a better lens for managing software delivery and cultural change, especially as artificial intelligence transforms the workplace.
What Systems Thinking Reveals#
Organizations aren't machines with simple cause-and-effect relationships. Instead, they operate as complex adaptive systems where any intervention ripples outward in unpredictable ways. Hendrickson describes it this way: "You poke the system, the system is going to poke back." Software delivery depends on layers of interlocking parts—people, technology, infrastructure, and open-source dependencies—all influencing one another simultaneously.
The interconnected nature means that what works in one context may backfire in another. The traditional linear thinking that worked for past transformations (like agile adoptions) falls short when organizations deploy AI across the board. Many teams are discovering that while AI accelerates output, the rest of the system isn't prepared to handle the volume. As Hendrickson notes, "AI is an amplifier. It's amplifying one part of the system more than the other parts of the system," exposing bottlenecks that didn't exist before.
A striking example: one study found that organizations adopting AI coding tools saw roughly 300% more code generated and 400% more bugs. Systems thinking explains why: the bottleneck simply shifted elsewhere.
Two Levers for Cultural Change#
Intentional cultural shift comes through two distinct mechanisms. Policy levers require leadership authority—hiring criteria, job descriptions, organizational structure. But cultural levers work differently: nobody can mandate culture directly. Instead, leaders and team members alike shape it by what they choose to feed.
Feeding culture means spotlighting what you want to grow. When a colleague submits a clean, well-organized pull request, specific praise ("that was so tight and easy to review") reinforces collaborative quality. When a team member says "I'd like to work with Elisabeth today because I like to see what she sees," they're feeding a culture of learning together rather than waiting until problems escalate.
These daily choices—how feedback is given, how questions are reframed, how demos tell stories rather than just display status—are available to everyone, not just executives. As Tosi emphasizes, small nudges compound over time.
