{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiasxf4znaev3i4t4czp6kzxyi67ip2xethyhaza4lbyyn4me2zgjy",
    "uri": "at://did:plc:qxttbtos2hod3osvlormrep7/app.bsky.feed.post/3mbekajeojt32"
  },
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreichgyuke2b24wwvv6zyna45kxq6xkt4aec64subusblc2fvggoghe"
    },
    "mimeType": "image/webp",
    "size": 89726
  },
  "description": "AI can speed up change work, but it can also quietly erode the thinking behind it. New research suggests overreliance on AI creates “cognitive debt,” weakening judgment, recall, and sensemaking if we’re not intentional about how we use it.",
  "path": "/ai-cognitive-debt/",
  "publishedAt": "2026-01-01T14:02:54.000Z",
  "site": "https://www.changeguild.co",
  "tags": [
    "recent research paper",
    "Let's Talk",
    "Support the Work"
  ],
  "textContent": "> **TL;DR:** A new study suggests that heavy reliance on AI for writing reduces cognitive engagement and memory retention. For change practitioners, this matters. AI can accelerate work, but if misused, it quietly erodes judgment, synthesis, and ownership—the very skills our profession depends on. The solution isn’t using less AI. It’s using it differently.\n\n## The Convenience Trap\n\nA recent research paper made waves in AI and academic circles for an uncomfortable reason.\n\nResearchers asked participants to write short essays under three conditions:\n\n  * with no assistance\n  * using a search engine\n  * using a large language model (LLM)\n\n\n\nThey didn’t just measure output quality. They measured _brain activity_.\n\nThe finding was striking:\nParticipants who relied on AI showed **lower cognitive engagement** , weaker recall of what they wrote, and less evidence of deep processing. The researchers called this accumulation of **“cognitive debt.”**\n\nNot burnout.\nNot laziness.\nDebt.\n\nWork got done. But understanding didn’t accumulate.\n\nIf you work in change, transformation, or strategy, that should make you uncomfortable.\n\nSubscribe\n\n## Why This Matters for Change Practitioners\n\nChange work is not mechanical work.\n\nIt relies on:\n\n  * sensemaking\n  * pattern recognition\n  * contextual judgment\n  * narrative construction\n  * political and emotional intelligence\n\n\n\nThese are _thinking-first_ skills.\n\nAnd they are exactly the skills most vulnerable to degradation when AI is used as a shortcut instead of a partner.\n\nThe risk isn’t that AI will replace practitioners.\n\nThe risk is that practitioners will slowly lose the very cognitive muscle that makes them valuable.\n\n## Cognitive Debt, Explained Simply\n\nThink of cognitive debt the way we think of technical debt.\n\nYou can:\n\n  * move faster now\n  * skip some hard thinking\n  * let a tool fill in the gaps\n\n\n\nBut later, you pay for it:\n\n  * weaker intuition\n  * poorer recall\n  * less confidence defending your work\n  * reduced ability to improvise under pressure\n\n\n\nOver time, this compounds.\n\nThe research suggests that when people _start_ with AI rather than _finish_ with it, their brains engage less deeply with the problem itself.\n\nAnd in change work, the problem _is the work_.\n\n## Where Practitioners Are Most at Risk\n\nBased on how most people are using AI today, the danger zones are predictable.\n\n### 1. First-Draft Thinking\n\nLetting AI generate:\n\n  * stakeholder analyses\n  * change narratives\n  * comms strategies\n  * impact assessments\n\n\n\n…before you’ve done your own synthesis.\n\nThis feels efficient.\nIt is also where the deepest cognitive outsourcing happens.\n\n### 2. Pattern Substitution\n\nAI is very good at producing _plausible patterns_.\n\nIt is not good at knowing:\n\n  * what matters politically\n  * what’s sensitive culturally\n  * what has already failed in this organization\n\n\n\nThose insights come from lived experience and judgment, not prompts.\n\n### 3. False Fluency\n\nWhen AI writes cleanly, people often mistake polish for clarity.\n\nBut clarity without comprehension is fragile.\n\nYou see it later when:\n\n  * you can’t explain your own deck\n  * your logic falls apart under questioning\n  * stakeholders push back and you don’t know why\n\n\n\n💡\n\n****Want support applying these ideas to your practice or team?**** We offer coaching to help change leaders do better work. Let's Talk.\n\n## The Right Way to Use AI in Change Work\n\nThe answer is not “use AI less.”\n\nIt’s **use AI later**.\n\nHere’s a model that works.\n\n### Step 1: Think First (Analog Brain On)\n\nBefore touching AI:\n\n  * Write the ugly version\n  * Sketch the logic\n  * Bullet your assumptions\n  * Note what you _don’t_ understand yet\n\n\n\nThis is where cognition happens.\n\nMessy is good.\n\n### Step 2: Use AI as a Challenger, Not an Author\n\nNow bring in AI to:\n\n  * pressure-test your thinking\n  * identify gaps\n  * suggest alternative framings\n  * surface counterarguments\n\n\n\nPrompt it like a sparring partner, not a ghostwriter.\n\nExample:\n\n> “Here’s my draft thinking. Where is this weak, incomplete, or overly simplistic?”\n\n### Step 3: Reclaim Ownership\n\nBefore anything goes out:\n\n  * rewrite in your own voice\n  * re-sequence the logic\n  * remove anything you don’t fully understand\n  * ask: _Could I explain this without notes?_\n\n\n\nIf the answer is no, you’re borrowing intelligence instead of building it.\n\n## A Simple Rule of Thumb\n\n> If AI saves you thinking time, you’re probably doing it wrong.\n> If AI sharpens your thinking, you’re doing it right.\n\nThe goal is not speed.\nThe goal is _better judgment per unit of effort_.\n\n## What This Means for the Future of the Profession\n\nChange practitioners who thrive in the AI era will:\n\n  * Think before they prompt\n  * Use AI to stress-test, not substitute\n  * Build personal frameworks instead of borrowing generic ones\n  * Maintain strong narrative and synthesis skills\n  * Treat cognition as a professional asset, not a cost center\n\n\n\nThose who don’t may still produce deliverables—but they’ll struggle when things get messy, political, or ambiguous.\n\nAnd change work is always messy.\n\n## Final Thought\n\nAI is not making us less capable.\n\nBut it _is_ revealing who was thinking deeply to begin with.\n\nIf you want to future-proof your practice, the answer isn’t resisting AI. It’s protecting the part of your work that only a thinking human can do.\n\n> **ChangeGuild** : Power to the Practitioner™\n\nJoin The Guild\n\n* * *\n\n### Now What?\n\n  * **Think before you prompt.**\nEven a rough outline or a few handwritten notes forces your brain to engage with the problem. If you start with AI, you skip the cognitive work that builds judgment. Use the tool after you’ve formed a point of view, not instead of one.\n  * **Use AI as a challenger, not a creator.**\nThe most valuable prompts ask what you missed, what assumptions you’re making, or how your thinking could fail. When AI generates the work for you, learning stops. When it pushes back on your thinking, learning accelerates.\n  * **Pressure-test your own understanding.**\nBefore sharing anything AI-assisted, ask yourself if you could explain it clearly without notes. If you can’t, the insight isn’t yours yet. That’s a signal to slow down, not ship faster.\n  * **Practice unaided thinking on purpose.**\nSet aside time where AI is off and thinking is manual. Write badly. Sketch ideas. Talk through problems. This isn’t inefficiency—it’s maintenance. Cognitive strength fades when it isn’t exercised.\n  * **Protect what actually makes you valuable.**\nIn change work, value comes from judgment, framing, and sensemaking—not speed or volume. AI can amplify those skills, but only if you keep them sharp. Otherwise, it quietly replaces them.\n\n\n\n* * *\n\n### Frequently Asked Questions\n\n**What is “cognitive debt” in the context of AI?**\nCognitive debt refers to the mental cost of repeatedly outsourcing thinking to tools like AI. When people rely on AI to generate ideas, structure arguments, or do initial reasoning, they often retain less understanding and build weaker mental models over time. The work gets done, but the learning doesn’t accumulate.\n\n**Is using AI making people less intelligent?**\nNot inherently. The research suggests the issue isn’t AI itself, but _how_ it’s used. When AI replaces early-stage thinking, it reduces engagement and recall. When it’s used to challenge or refine thinking, it can actually deepen understanding.\n\n**Should change practitioners stop using AI altogether?**\nNo. That would be unrealistic and counterproductive. The goal is not avoidance, but intentional use. AI works best as a thinking partner, not a substitute for judgment, sensemaking, or experience.\n\n**What kinds of tasks are most at risk of cognitive debt?**\nTasks that involve synthesis, framing, or decision-making are the most vulnerable. Examples include stakeholder analysis, change narratives, strategy documents, and executive communications. These require understanding, not just output.\n\n**How can I tell if I’m relying too much on AI?**\nA simple test: if you struggle to explain or defend the work without looking at it, you probably didn’t internalize the thinking. That’s a sign the tool did more of the work than you did.\n\n**Does this mean AI reduces productivity in the long run?**\nNot necessarily. AI can dramatically improve efficiency. The risk is when efficiency replaces comprehension. The most effective practitioners use AI to accelerate insight, not bypass it.\n\n**What’s the safest way to use AI in change work?**\nStart with your own thinking. Use AI to test, challenge, or expand it. Then take ownership of the final output. If you can explain it clearly without the tool, you’re using AI the right way.\n\n* * *\n\n💡\n\n****Like what you’re reading?****\nThis post is free, and if it supported your work, feel free to support mine. Every bit helps keep the ideas flowing—and the practitioners powered. [Support the Work]",
  "title": "How to Use AI Without Losing Your Edge",
  "updatedAt": "2026-06-05T01:27:22.067Z"
}