The Amplification Toolkit

10 min read | Topic: How?

Most advice about AI focuses on what it can do for you. This chapter flips that frame: what can AI help you do more of?

The professionals who will command premium value aren’t racing AI to the bottom—competing on speed and volume in tasks AI already does well. They’re racing to the top—identifying what makes them uniquely valuable and using AI to multiply that value across more situations, more people, more impact.

This isn’t about finding work AI can’t do (that list shrinks daily). It’s about finding work that you do distinctively well, then building systems that scale your distinctive contribution.

Layer 1: Expertise Mapping

Before you can scale your capabilities, you need to know what they actually are. This sounds obvious. It isn’t. Most professionals have never systematically mapped their unique value—they just do their jobs and hope someone notices.

Identifying Your “Unreplaceable 20%”

Not all of your work is equally valuable or equally yours. The Pareto principle applies: roughly 20% of what you do likely generates 80% of your distinctive impact. Your job is to find that 20%.

The Replacement Thought Experiment

Imagine your role was filled by a capable AI agent with access to all your documents, emails, and standard procedures. What would go wrong? Not “what tasks wouldn’t get done”—AI can do most tasks. But what outcomes would suffer? What relationships would fray? What quality would decline?

The gaps you identify are clues to your unreplaceable contribution.

The “They Call Me When…” Audit

Think about the last ten times a colleague, client, or friend specifically sought your input. Not just anyone who could help—you. What was the pattern?

What problems do people bring to you that they don’t bring to others?
What decisions do people want your read on?
What situations make people say “we need to get [your name] involved”?

These patterns reveal your perceived unique value. Take them seriously—others often see our strengths more clearly than we do.

The Energy Audit

Track your energy across a typical week. When do you feel most alive, most engaged, most like yourself? When does time disappear? These aren’t just preference signals—they’re competence signals. We typically have energy for work we’re genuinely good at.

Conversely, notice when you feel drained or fraudulent. That work might be important, but it’s probably not your 20%.

The Skills AI Makes More Valuable, Not Less

Here’s a counterintuitive truth: some human skills become more valuable as AI improves, not less. Understanding this changes how you invest in your own development.

Judgment in Novel Situations

AI excels at pattern-matching against training data. Humans excel at reasoning through genuinely new situations—especially those requiring integration of context AI doesn’t have access to. As AI handles more routine decisions, the remaining decisions are increasingly non-routine. Judgment becomes scarcer, therefore more valuable.

Relational Trust

AI can simulate warmth. It cannot build trust earned through shared history, demonstrated reliability, and genuine mutual investment. As AI mediates more interactions, unmediated human relationships become rarer and more precious. The professional who can build genuine trust has an asset AI cannot replicate.

Taste and Curation

When anyone can generate infinite content, the ability to discern what’s good becomes the bottleneck. Taste—the cultivated capacity to recognize quality, relevance, and resonance—is a human skill that AI amplifies rather than replaces. The curator becomes more valuable than the creator.

Accountability and Ownership

Someone has to stand behind the work. Someone has to be responsible when things go wrong. Someone has to care about outcomes beyond the immediate task. AI can produce; humans must own. This accountability—the willingness to stake reputation on results—is a form of value creation AI cannot provide.

Personal API: What You Offer That AI Cannot

Think of yourself as a service with a defined interface. Your “Personal API” is the set of capabilities others can reliably call on you for—your documented, dependable value proposition.

Defining Your Endpoints

What are the specific “functions” you offer? Be concrete:

“I can look at a business model and identify the three things most likely to kill it”
“I can take a room full of disagreeing stakeholders and find the decision everyone can live with”
“I can translate technical complexity into language executives actually understand”
“I can sense when a team is about to break down before anyone says anything”

These aren’t job descriptions. They’re capabilities—the atomic units of value you provide.

Documenting Your Parameters

Good APIs have clear documentation. What inputs do you need to do your best work? What context? What access? What timeline? Understanding your own requirements helps you (and others) deploy your capabilities effectively.

Understanding Your Limitations

Honest APIs document what they don’t do. What are you mediocre at? What do you actively avoid? What situations bring out your worst? This isn’t self-criticism—it’s useful information for directing your energy toward your 20%.

Layer 2: Multiplication Practices

Once you’ve mapped your unique capabilities, the next question is: how do you scale them? AI enables forms of multiplication that weren’t previously possible.

One-to-Many Frameworks via AI

Your expertise currently scales linearly—you can only be in one meeting, one conversation, one project at a time. AI can help you scale non-linearly.

Codifying Your Thinking

Can you teach AI to think like you about specific problems? Not generally—AI won’t become you—but narrowly, for defined situations?

Try this: Take a type of decision you make repeatedly. Document your actual reasoning process in detail. What do you look for? What patterns concern you? What questions do you always ask? What trade-offs do you weigh? Feed this to AI as a framework, then test it against past decisions. Iterate until it approximates your judgment.

The result isn’t a replacement for you—it’s a first-pass filter that handles routine cases and escalates edge cases to your actual attention.

Creating Decision Trees from Your Intuition

Much of expert judgment feels intuitive but actually follows implicit logic. Work with AI to make that logic explicit:

Describe a recent decision you made
Have AI ask you “why” questions until you’ve articulated your reasoning
Generalize the reasoning into principles
Test the principles against other decisions

This process extracts the structure hidden in your intuition—structure that can then inform AI tools, junior colleagues, or documentation.

Building Your Knowledge Base

Create a persistent repository of your expertise that AI can draw on:

Your frameworks and mental models
Your past analyses and recommendations
Your principles and heuristics
Your vocabulary and definitions

This isn’t just for AI—it’s for future you, and for anyone who needs to understand how you think.

Async Collaboration with AI Agents

You don’t have to be present for AI to extend your capabilities. Design workflows where AI agents carry your judgment into contexts you can’t personally attend.

The Briefed Agent Model

Before important work happens without you, brief an AI the way you’d brief a trusted deputy:

Here’s what I care about
Here’s what I’d be watching for
Here’s what would concern me
Here’s when to escalate

Then have the AI monitor, filter, or pre-process according to your criteria. You’re not delegating judgment—you’re extending your attention.

Asynchronous Review Cycles

Instead of synchronous collaboration (you and AI in real-time), try asynchronous cycles:

Set up a task with your criteria and constraints
Let AI work independently
Review results in batch
Provide feedback that improves the next cycle

This lets you maintain quality control while dramatically increasing throughput.

The “Clone Your Judgment” Exercise

This exercise stress-tests how well you understand your own expertise.

Step 1: Select a Domain

Choose an area where you have genuine expertise and make judgment calls regularly.

Step 2: Teach AI to Be You

Create a comprehensive prompt that would allow AI to make decisions the way you would in this domain. Include your priorities, your heuristics, your red flags, your exceptions. Be exhaustive.

Step 3: Test Against Reality

Present the AI with past decisions you’ve made (without revealing your actual choice). How often does it match your judgment? Where does it diverge?

Step 4: Analyze the Gaps

The gaps are gold. They reveal either:

Tacit knowledge you haven’t articulated (make it explicit)
Inconsistencies in your own judgment (examine them)
Contextual factors AI can’t access (document them)

Even if you never use the resulting “clone,” the exercise forces a level of self-understanding most professionals never achieve.

Layer 3: Creative Combination

The highest form of human-AI collaboration isn’t automation or augmentation—it’s combination. Using AI to explore possibilities that neither human nor machine would reach alone.

Human Intuition + AI Exploration

Your intuition is a powerful but narrow beam. AI can illuminate the vast space around it.

The “Expand Then Select” Method

Start with your intuition: What direction feels right?
Use AI to generate variations: What are twenty other directions in this space?
Apply intuition again: Which of these resonate? Which combinations spark something?
Iterate: Expand around the promising areas

Your intuition guides exploration. AI handles the exploration. Your intuition judges the results. This cycle covers far more ground than either could alone.

The “Why Not?” Protocol

For any decision or creative direction, ask AI: “What are all the reasons this might be wrong, or alternatives I might not have considered?” Use AI as a systematic devil’s advocate against your own intuitions.

You’ll often confirm your original instinct. But occasionally, you’ll discover blind spots that change your direction. That’s the value.

Taste-Making in the Age of Infinite Generation

When generation is free, curation becomes the scarce resource. Your taste—cultivated over a lifetime—is a competitive advantage.

The Curator’s Workflow

Generate abundantly (use AI to create far more options than you need)
Curate ruthlessly (apply your taste to select the 5% worth developing)
Refine personally (add the human polish that elevates good to excellent)

Most people generate conservatively and accept too much. Invert that: generate liberally and accept almost nothing.

Developing Articulable Taste

Taste that can’t be explained can’t be scaled. Work on articulating why you prefer what you prefer:

What specifically makes this better than that?
What principle does this example embody?
What would I change to make this good enough?

The more precisely you can articulate taste, the more effectively you can direct AI toward what you actually want.

The “Adjacent Possible” Method

Stuart Kauffman’s concept of the “adjacent possible” describes the set of things that are one step away from current reality—not yet existing, but reachable. AI dramatically expands your adjacent possible.

Mapping Your Edges

What are the boundaries of your current work? The edges of your expertise? The limits of what you’ve tried? These edges are where adjacent possibilities live.

Systematic Edge Exploration

Use AI to explore just beyond each edge:

“What’s a version of this that’s slightly more ambitious?”
“What would this look like in an adjacent field?”
“What’s an assumption I’m making that I could question?”
“What would someone with a completely different background try here?”

You’re not looking for AI’s ideas. You’re using AI to map territory that your human intuition can then evaluate.

Combination as Creation

The most powerful adjacent possibilities often come from combining existing elements in new ways. Use AI to generate combinations you wouldn’t naturally consider:

Your expertise + unfamiliar methods
Your industry + adjacent industry practices
Your approach + contrasting approaches

Most combinations will be worthless. But the valuable ones—the ones that make you think “huh, that’s interesting”—are seeds that only human cultivation can grow.

The Scaling Mindset

The throughline of this chapter is a shift in how you think about AI’s role in your work:

Old mindset: “What can AI do for me?”
New mindset: “What can AI help me do more of?”

The first mindset leads to delegation and replacement. The second leads to amplification and multiplication. The first treats you as a bottleneck to be routed around. The second treats you as a source of value to be scaled.

Start by mapping your 20%—the work that’s distinctly yours. Then systematically build the infrastructure to multiply it. The professionals who do this won’t be competing with AI. They’ll be competing with other humans who have AI as their multiplier.

Make sure your multiplier is pointed at something worth multiplying.