In one line: Treat AI as a capable teammate in a supporting role. Use it to extend what people can do, but keep a human in the lead and accountable for the result.
Core idea
AI tools now draft, summarize, search, code and analyze. The leadership question is no longer whether teams will use them but how. The program’s stance is simple: the human is in the lead. AI can speed up work and widen the options a team considers, but judgment, responsibility and ethical accountability stay with people.
How it works
- Know the spectrum. It runs from narrow predictive models through generative AI (which creates new text, images or code resembling its training data) to agentic systems that take actions. Each step up adds capability and risk.
- Start from use cases, not tools. Find tasks where speed or breadth matters and errors are catchable, such as first drafts, literature scans, test-case generation or data cleanup.
- Govern deliberately. Decide what data may be shared, how outputs are checked, who signs off, and how use is disclosed.
- Build trust carefully. Models can be confidently wrong. Verification effort should scale with the consequences of an error.
- Redesign the workflow. The biggest gains come from rethinking who does what, not from bolting a chatbot onto an old process.
In practice
For an engineer, sensible patterns include using AI to draft documentation or scripts that are then reviewed, to brainstorm design alternatives that are then evaluated with engineering judgment, or to search a large body of reports for relevant prior work. Less sensible: accepting AI-generated analysis into a design review without independent checks, or putting controlled or proprietary data into tools that are not approved for it. Courses increasingly ask for a statement of how AI was used, and that habit of disclosure is worth carrying into work.
Questions to ask yourself
- Which of my recurring tasks could AI speed up without increasing risk?
- How would I catch an AI error before it reached a customer or a design review?
- Could I explain and defend every AI-assisted result I sign?
Connections
- Extends: Teaming to non-human teammates.
- Engineering side: Software in Engineered Products. Risk side: Risk Management.
- Organizational side: Adaptive and Transformational Change.
- Capabilities exercised: Ethical Actions and Integrity, Decision Making, Resourcefulness.
Sources
- Course podcast and guest discussion on AI and teaming; no public source.