Lisa Andrews on AI Automation and Leadership

I've been collaborating with Claude (Anthropic's AI) on a complex creative project — roughly 15 structured work sessions over several weeks involving interconnected systems design, architectural decisions, and large-scale document management. At one point, I asked it for an honest, non-flattering assessment of my working style and contributions. Not a performance review. Not a summary of what I'd done. An observation of how I work, from something that had watched me work closely for weeks. What came back was more precisely observed than most 360 feedback I've received from humans. A few highlights: "His primary mode is structural decision-making: he identifies the constraints that matter, makes calls quickly, and rarely reverses them. When he does reverse a decision, it's because new information changed the problem, not because the original judgment was unsound." "He holds complex interdependent systems in working memory... checking proposals against established constraints, mechanical systems, experience implications, and coherence across multiple workstreams. He does this without external reference most of the time — the architecture lives in his head and he navigates it fluently." "He provides seed ideas — often a single sentence — and evaluates what comes back against an internal standard that accounts for the full system simultaneously." On how I use AI specifically: "He understands what the tool is good at (volume generation, consistency tracking, detail expansion, finding connective tissue between established decisions) and what it isn't (making the structural calls, evaluating quality, holding the full vision). He directs accordingly." I'm sharing this because I think we're past the point of debating whether AI is useful. The more interesting question is how leaders work with it. In my experience, the highest leverage comes from treating AI as a force multiplier for architectural thinking — not a replacement for judgment, and not a fancy autocomplete. You bring the vision and the decisions. It brings volume, consistency, and connective tissue. The skills that make someone effective at this aren't new: systems thinking, precise communication, knowing when to lock a decision and when to leave it open. What's new is that those skills now compound differently. #AICollaboration #SystemsThinking #TechnologyLeadership #FutureOfWork #ExecutiveLeadership

Most leadership teams made the same call in the last two years. Junior roles doing work that AI can now do faster. Costs down, velocity up. On paper, a clean trade. I've been thinking about what that trade actually bought. In the 1980s and 90s, the aviation industry automated the cockpit. Crashes went down. The data was unambiguous. What wasn't visible was what was eroding underneath — the manual skills, the situational awareness, the judgment you only build from thousands of hours of doing the work yourself. By 2009 researchers had a name for it: automation dependency. 77 percent of commercial pilots surveyed reported their skills had deteriorated. The industry didn't discover the problem when they made the call to automate. They discovered it years later, when the system failed and nobody knew what to do. I think most organizations are in the early innings of the same pattern. Junior positions aren't primarily about output. They're about absorption — learning how decisions actually get made, which constraints are real, what good costs in this specific org with these specific people. A 2025 MIT Sloan study found that the organizations getting the most from AI aren't using it to get answers faster. They're using it to design better decision environments. That requires people who know which questions are worth asking. You build that by doing the work at every level. You don't build it by watching AI do it. The competence lag is what makes this hard to catch. Your org looks fine right now. The problems show up in three to five years. Full piece in The Gap — link in bio.

Reminder: Early registration savings for AI and Business Strategy ends March 30. AI is rapidly becoming a core component of business strategy. Leaders who understand how to strategically apply AI will be better positioned to guide their organizations through this transformation. Starting April 28, this WatSPEED course will help you: • Evaluate where AI can create the most value • Align AI initiatives with business objectives • Lead AI-driven innovation within your organization Our AI and Business Strategy course has had more than 500 learners from more than 180 organizations since its inception. The course has had graduates who work in organizations like Health Canada, OpenText, CIBC, Canada Life, and Waterloo Regional Health Network. Register before March 30 and use code EARLY10 to receive 10% off enrolment. https://lnkd.in/eM8TiWSx #AI #Business #Education

The Skill That Becomes Priceless When AI Takes Over It’s Not Creativity. It’s the Willingness to Be Wrong. I’ve been in consulting for 40 years. The single biggest predictor of whether a project succeeds or fails has nothing to do with strategy, budget, or technology. It’s whether one person in the room is willing to say, “This is my decision and I own the outcome.” Not a committee. Not a task force. Not a consensus-driven process that distributes accountability so evenly that nobody is actually accountable for anything. One person. One decision. One name attached to it. The organizations that move fastest are the ones where someone raises their hand before they’re asked. The ones that stall are the ones where everyone waits for someone else to go first. Peter Drucker put it simply: “Wherever you see a successful business, someone once made a courageous decision.” In 2026, with AI generating options by the thousands and data available by the terabyte, the bottleneck isn’t information. It’s the willingness to act on it. The skill that becomes priceless isn’t analysis. It’s judgment. And judgment requires someone willing to be wrong. Who’s that person in your organization? Be well, Rod #Leadership #DrRodWilson #BeneathTheSurface #HumanIntelligenceEra

🧠 Most people think Prompt Engineering is just "asking AI better questions." It's not. And as tech leaders, this misunderstanding is costing our teams more than we realize. 𝗪𝗵𝗮𝘁 𝗣𝗿𝗼𝗺𝗽𝘁 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴 𝗿𝗲𝗮𝗹𝗹𝘆 𝗶𝘀: Prompt Engineering is context architecture. It's the discipline of structuring: → Role (who the AI should be) → Task (what exactly needs to be done) → Constraints (what to avoid or prioritize) → Format (how the output should look) → Examples (what "good" actually means) When all five are aligned, AI doesn't just respond — it performs. 𝗙𝗮𝗰𝘁 #𝟭: It's not about wording. It's about structure. A well-engineered prompt is closer to an API contract than a search query. The AI is not reading your mind — it's reading your context. Give it a poor context, and even the most powerful model will produce mediocre output. 𝗙𝗮𝗰𝘁 #𝟮: A bad prompt is not just annoying — it's expensive. Here's what poor prompting actually costs in a production engineering environment: ⚠️ Hallucinations that slip into code reviews or documentation ⚠️ Rework cycles as teams iterate on vague AI outputs ⚠️ Brittle AI features that behave differently across inputs ⚠️ Eroded trust — developers stop relying on AI tools altogether In fast-moving teams, these aren't minor inconveniences. They compound into sprint delays, quality debt, and missed opportunities. As leaders, we don't need to write every prompt ourselves. But we DO need to understand what makes a prompt effective — so we can set standards, enable our teams, and make smarter decisions about where and how AI fits into our workflows. Prompt Engineering is the new requirement gathering. The better you define the problem, the better the AI solves it. Follow along if you're on the same journey of becoming an AI-native engineering leader. 🚀 #PromptEngineering #GenerativeAI #AILeadership #TechLeadership #LLMs #EngineeringExcellence #AIStrategy #Upskilling #AINative #ContinuousLearning

AI is making it easier to build. But it is not making it easier to decide WHAT to build. Prototypes that used to take many weeks can now be created in a few hours. Ideas now move faster from concept to working artifacts. But building faster is only part of the equation. The more difficult question is whether organizations can absorb that speed. Can teams: * evaluate more ideas? * make decisions faster? * align across functions? * bring products to market effectively? In most cases, engineering is no longer the bottleneck. It is decision-making, coordination, and go-to-market readiness. Unless the speed of decision-making catches up with the speed of building outpaces the organizations don’t accelerate. AI changes the economics of execution. But unless operating models evolve as well, the system does not get faster. #AIProducts #ProductLeadership #ProductManagement #SystemsThinking #GTM

AI Is Reshaping Product Decisions, Not Just Product Development. “Why are we even building this with AI?” A colleague asked me this recently. It’s a fair question. Most of what we’re building with AI today could be built without it. But that’s the wrong lens. In product leadership, the constraint has never been the ability to build. It’s been the cost of being wrong. Every decision carries weight: engineering allocation, stakeholder alignment, and organizational momentum. Once a direction is funded, it becomes harder to reverse regardless of whether it’s right. AI doesn’t eliminate that risk. It changes when and how we take it. Recently, I explored a simple idea: a lightweight experience to help children process emotional overwhelm at the end of the day. In a traditional environment, this would have entered a backlog, competing for priority, waiting for validation. Instead, I used AI-assisted prototyping to translate the concept into a functional interaction within hours. Not production-ready. Not scalable. But sufficient to answer a more important question: Does this idea create meaningful user signal? That is the shift. AI compresses the distance between assumption and evidence. This has three implications for product organizations. First, exploration and execution can now be decoupled. We no longer need to commit full engineering resources to validate direction. Ideas can be explored quickly, and only the strongest signals are scaled. Second, the unit of progress becomes learning not output. Shipping features is no longer the primary indicator of progress. Reducing uncertainty is. Third, product leadership becomes more accountable, not less. When it is easier to test ideas, there is less justification for investing in unvalidated assumptions. Judgment - not execution becomes the differentiator. This does not replace engineering discipline. It does not change the need for scalable architectures. And it does not make technology decisions irrelevant. But it does change the sequence. Clarity before scale. Organizations that treat AI as a productivity tool will move faster. Organizations that treat AI as a decision tool will move smarter. The question is no longer: Can we build this? It is: How quickly can we prove ourselves wrong? Curious how others are thinking about this - are you optimizing for output, or for learning? #vibecoding #productleadership #AI #DigitalTransformation #DecisionMaking #LeanProduct #Experimentation

If you're exploring ways to use AI in your business, take a look at this. DeVryPro offers a FREE AI for Entrepreneurs course that shows how AI tools can help validate ideas, prototype early assets, and automate operations. From there, you can continue learning with courses in AI leadership, self-leadership, and product management. Check it out: https://devry.ly/4bOZiQa #AIForEntrepreneurs #AIAtDeVry #DeVryPro #TeamDeVry https://devry.ly/481H78N

If you're exploring ways to use AI in your business, take a look at this. DeVryPro offers a FREE AI for Entrepreneurs course that shows how AI tools can help validate ideas, prototype early assets, and automate operations. From there, you can continue learning with courses in AI leadership, self-leadership, and product management. Check it out: https://devry.ly/4bOZiQa #AIForEntrepreneurs #AIAtDeVry #DeVryPro #TeamDeVry https://devry.ly/4tQ5Oxx

If you're exploring ways to use AI in your business, take a look at this. DeVryPro offers a FREE AI for Entrepreneurs course that shows how AI tools can help validate ideas, prototype early assets, and automate operations. From there, you can continue learning with courses in AI leadership, self-leadership, and product management. Check it out: https://devry.ly/4bOZiQa #AIForEntrepreneurs #AIAtDeVry #DeVryPro #TeamDeVry https://devry.ly/4c7Wkrk