John Cutler's Post
John Cutler · article · 2026-07-31 Reader 🖥️
AI Confidence TheaterEnterprise AI Adoption & ReadinessFeedback LoopsLearning Culture & SlackMeta-Skills & Contextual JudgmentStrategy & Direction-SettingTrust
In an environment where every competitor has access to the same AI technology, durable advantage comes not from adopting AI fastest but from building — through problem-first, hypothesis-driven experimentation and a genuine social contract of trust — a team practice that learns faster than the surrounding hype cycle.
practical · mixed — Arrives in Reader looking like a ~56-word LinkedIn teaser ('Critical skill: learning faster than the hype... might as well share!'), but the document's content field embeds all 34 images of the underlying presentation deck ('Learning Faster Than the Hype (Together)') that Cutler built for his Dotwork client-facing talks. This extraction reads that deck directly (downloaded and visually read each slide image). It is genuinely substantive — not a thin teaser — but it is a slide deck, not connected prose: expect bullet-point density, recurring diagrams, and inferred connective reasoning between slides rather than an explicit written argument.
Structure
| Part | Role |
|---|---|
| Opening reframe: 'You are not crazy' / the Just...But...Can triangle | context |
| Why competition is no longer about the technology itself (Q: how do you compete when everyone has the same tech? A: judgement, execution, customer understanding, trust, ability to learn/adapt; the sledgehammer-between-shack-and-mansion sketch) | evidence |
| Competition-and-Clarity 2x2 and the definition of 'strategic clarity' as a grounded thesis (not certainty) | framework |
| Maturity map of current AI practices ('emergent' not yet 'established') and the four-point durable-advantage checklist (act without a playbook, learn amid the hype, filter advice, convert learning into advantage) | framework |
| Three case stories (escaping the AI build trap; 'green but not green' code migration; OKR alignment from 7/10 to 10/10), each with a before/after and what-they-tried/what-they-noticed table | evidence (narrative) |
| Success formula: Success = (Machine Understanding × Problem Understanding × Practice Evolution) / Social Contract, plus 'where current AI is effective' criteria and the verifiability/context/feedback/trust comparison table | synthesis / framework |
| Three Doors (machine-first, problem-first, practice-first) with their attached blind-spot questions, and a worked hypothesis example | framework / method |
| 'Avoid the traps' — eight recurring organizational failure patterns AI can help escape or deepen; closing 'six months from now...' prompt | conclusion / call to reflection |
Key terms
- Learning faster than the hype
Building a team/organizational learning velocity — real signal from customers, systems, and colleagues — that outpaces the churn of AI narratives, vendor claims, and 'advice firehose', so that adoption is grounded rather than reactive. - Just... / But... / Can
Shorthand for the two unproductive default postures in AI debates — advocates who say 'just use AI / just adopt it' and skeptics who counter 'but what about quality/context/people' — reframed into the productive question 'what CAN we do?' - Three Doors
The three legitimate entry points into an AI initiative — machine-first ('AI can do X'), problem-first ('we have a problem'), practice-first ('teams are doing Z') — each legitimate but each carrying its own blind-spot questions the initiative must still answer. - Strategic clarity
Not pretending to have all the answers; a grounded thesis about what matters, what remains uncertain, and how you will learn — distinct from confident-sounding certainty.
Propositions
- explicitLearning faster than the hype is the critical differentiating skill in the current AI moment.conf: high
- explicitWhen everyone has access to the same AI technology, competitive advantage comes from judgement, execution, customer understanding, integration, design, distribution, trust, and the ability to learn and adapt — not from the technology itself.conf: highsupportsTBM 420: The AI Playbook Puzzle · p5 — Naming judgement, execution, and adaptability as the real differentiators when AI access is equal supports his earlier claim that systems thinking, metacognition, and contextual judgment are being discounted precisely when they're most needed.
- implicitThe value AI unlocks is inseparable from who wields it and the surrounding organizational context — the same capability produces a rebuilt house or a ruin depending on the builder.conf: mediumsupportsAI Won't Save the Kingdoms We Built · p7 — The sledgehammer-between-shack-and-mansion sketch (the tool's value depends on the wielder/context) supports his earlier, more explicit claim that AI is only genuinely useful as an amplifier of an already-coherent operating model and cannot substitute for one.
- explicitHigh competitive pressure paired with low strategic clarity produces FOMO-driven activity (mandates, scattered pilots, tool buying, AI theater) rather than real progress.conf: high
- explicitStrategic clarity is not pretending to have all the answers; it is a grounded thesis about what matters, what remains uncertain, and how you will learn.conf: high
- explicitMost current AI practices (prompt engineering, agents, RAG, context engineering, vibe coding, evals) are still emergent and variable rather than stable and codified.conf: high
- explicitDurable advantage requires acting without a playbook, learning amid the hype, discerning which advice to heed and which to ignore, and converting what is learned into durable advantage.conf: highextendsTBM 420: The AI Playbook Puzzle · p8 — Cutler's 'act without a playbook' point in this deck extends his own earlier claim that nobody currently has a valid AI playbook, regardless of how confidently they market one.
- implicitTeams that shift from using AI to accelerate existing (often broken) practices toward using AI to increase real signal from users, systems, and each other get better outcomes even at short-term cost.conf: high
- explicitSuccess with AI = (Machine Understanding × Problem Understanding × Practice Evolution) / Social Contract.conf: high
- explicitThe social contract is not a nice-to-have; it is the precondition for honest experimentation — without it, you get compliance theater.conf: highsupportsPlease stop the AI Confidence Theater · p7 — Cutler's 'social contract as precondition for honest experimentation, else compliance theater' directly echoes Elena's framing of trust as the scarce, withdrawable shared resource of the AI era.
- explicitCurrent AI is most effective where verifiability, context accessibility, feedback density, semantic stability, source trustworthiness, and pattern availability are all high, and least effective where they are missing.conf: high
- explicitThere are three legitimate ways into an AI initiative — machine-first, problem-first, practice-first — and each carries a characteristic blind spot.conf: high
- explicitA well-formed AI experiment needs a specific hypothesis, a way to observe impact, a named stopping condition, and an explicit pivot-or-proceed decision.conf: high
- explicitdevelopedAI can help escape long-standing organizational failure patterns (the Build Trap, Premature Scaling, Limits to Growth, the Success Trap, Permanent Refinance, the Zombie Portfolio, the Commons Trap, Shifting the Burden) or deepen them, depending on how it's applied.conf: highextendsTBM 430: Incubate, Compound, Refinance, Liquidate · p3 — Extends Cutler's own Build Trap concept to a fuller list of eight organizational traps, with the same unresolved 'AI can help or deepen it' symmetric structure the corpus already flags as under-specified in his tbm-430 piece.extends ←Pascal BORNET's Post · p3 — Cutler's post names 'Shifting the Burden' as one item in a list of organizational failure patterns without elaborating it; Bornet's post is a compact, concrete walk-through of exactly that systems archetype — a symptomatic fix (workaround) that relieves the symptom and thereby crowds out the fundamental fix (redesign).
Arguments
Sustainable competitive advantage requires investing in a socially-trusted, hypothesis-driven learning practice, not simply adopting AI tools faster than competitors.
Explicit premises:
- Everyone has access to the same AI technology (p2's premise).
- The remaining differentiators are judgement, execution, customer understanding, trust, and the ability to learn and adapt (p2).
- Strategic clarity — not certainty, but a grounded thesis plus a plan to learn — beats FOMO-driven high-pressure activity (p4, p5).
- Most current AI practices are still emergent and unstable, not established best practice (p6).
Implicit premises:
- Technology alone does not determine the outcome — the same tool produces very different results depending on the wielder and context (p3, the sledgehammer sketch).
- An organization's practice and trust level mediate whether AI helps or hurts, as shown across all three case stories (p8, p14).
Inferences:
- Organizations should therefore budget at least as much attention to developing shared practice, judgement, and trust as to acquiring AI capability — a conclusion the deck implies via its case stories and success formula but never states as directly as this.
Honest experimentation with AI requires a genuine social contract; without it, results devolve into compliance theater.
Explicit premises:
- 'The social contract is not a nice-to-have. It is the precondition for honest experimentation. Without it, you get compliance theater.' (p10)
- The success formula places Social Contract in the denominator, so a weak or broken social contract degrades the whole calculation regardless of how strong the other three factors are (p9).
Implicit premises:
- People will not surface true negative results, dissent, or practice failures to a system that punishes honesty about them.
- Absent trust, teams will over-report success and under-report the friction and rework the case stories describe (echoed in the third story: 'real alignment is expensive').
Inferences:
- Leaders should treat trust-building as a prerequisite investment made before scaling an AI practice change, not an afterthought bolted on after tools are already deployed.
Narratives
- 'Escaping the AI build trap': a team used AI to produce more polished PRDs, better handoffs, and 2x+ PR throughput, and celebrated the speedup — but noticed less user interaction, conflated feedback, 'too much change,' and less learning signal. They pivoted to throwaway prototypes, involving engineers earlier, using AI to boost signal (not just speed), and having engineers rebuild the validated prototype. Outcome: fewer items shipped, but higher signal, more leverage, and more impact; engineers trusted to ship fewer, better bets.
→ Supports p8 and p14 — shows AI amplifying a broken build-first practice until the underlying practice itself changed toward customer contact and validated learning. - 'Green but not green': a team used AI for rapid code migration — a sophisticated pipeline, loop-until-green, file-by-file translation, done in a fraction of the time — but discovered low institutional knowledge transfer, more production bugs, and harder debugging, because 'code comprehension is not system comprehension.' They pivoted to focusing on intent, domain slicing, a portfolio approach, and a human-in-the-loop validation layer. Outcome: slower at first, no major regressions, engineers felt ownership, and the documentation produced became a lasting asset.
→ Supports p11 (AI pattern-matches surface syntax well but not semantic/system intent) and p6 (current AI practice around migration is still immature). - 'From 7/10 to 10/10': a team used AI as an OKR writer and 'cascader' — output looked coherent and leadership felt aligned, but the OKRs actually scored 7/10 despite the polish, conflicted with each other, and had 'no edge, all polish,' because real alignment is expensive. They pivoted to using AI-generated drafts only as the start of a conversation, requiring people to articulate priorities in their own words, and adding a back-briefing step. Outcome: the offsite took longer, but produced fewer and sharper OKRs, less auto-cascading, and less priority churn.
→ Supports p9 and p10 — shows that speeding up an artifact (the OKR document) with AI does not substitute for the social, effortful work (the social contract) that produces real alignment.
Questions addressed
- ✅ How do you compete when everyone has access to the same technology?
Judgement, execution, customer understanding, integration, design, distribution, trust, and the ability to learn and adapt — i.e., organizational and human factors, not the technology itself. - ⭕ How much time should I invest in the hyped approach of the day vs. the enduring foundations of my craft vs. going deep into how AI actually works?
- ⭕ What determines whether a given organizational 'trap' (Build Trap, Zombie Portfolio, Commons Trap, etc.) is helped or deepened by applying AI to it?
Gaps
- incompleteThe deck compresses three detailed organizational case studies (build-trap escape, code migration, OKR alignment) into a handful of two-to-four-word bullets per box ('Portfolio approach', 'Layer HIL validation', 'Domain slicing', 'Back briefing'), with no operational detail on how these practices were actually implemented step by step, who owned the change, or how long it took to land.
Read next: Cutler's longer TBM newsletter essays (already in this corpus, e.g. extract-tbm-430-incubate-compound-refinance-liquidate, extract-tbm-432-bundling-unbundling-capabilities) or a recording/write-up of the underlying live Dotwork client presentation would likely supply the missing mechanism. - incompleteThe 'Success = Machine Understanding x Problem Understanding x Practice Evolution / Social Contract' formula uses multiplication and division as if the four factors were measurable quantities, but none are scored, operationalized, or given units anywhere in the deck — it functions as a mnemonic device, not an actual computable model.
Read next: This is a recurring pattern in Cutler's own frameworks — the corpus already flags the same failure to operationalize a 2x2's axes in his tbm-430 piece (extract-tbm-430-incubate-compound-refinance-liquidate#g1) — worth reading together to see how consistently the gap recurs. - incompleteAll three case stories are presented as success narratives — the pivot worked in every case. The deck gives no counter-example of a team that adopted the same problem-first, hypothesis-driven, trust-first practice and still failed, nor any base rate for how often the 'before' pattern actually resolves the way these stories describe. Survivorship bias in the anecdotes is unaddressed.
Read next: Practitioner retrospectives or failure-mode write-ups of AI-adoption playbooks would help test whether the practice generalizes beyond hand-picked success cases. - incompleteThis LinkedIn post is a by-product of a live client talk ('Presenting to some of the biggest companies... I have a bad habit of creating new presentations and never doing the same one twice. So might as well share!'). The verbal narration, audience Q&A, and worked-exercise detail that would have accompanied these slides in the room are not captured here — only the static slide text and imagery survive as a 34-image carousel, so the connective reasoning between slides is partly inferred rather than stated.
Read next: A recording or write-up of one of the actual Dotwork client sessions, if it surfaces, would fill in the missing connective tissue.
Related documents
- AI Won't Save the Kingdoms We Built
- TBM 430: Incubate, Compound, Refinance, Liquidate
- AI Enthusiasts Are in a Race Against Time, AI Skeptics Are in a Race Against Entropy
- How to Build a Superteam That Keeps Getting Better
- Your AI strategy has a trust problem, not a tooling problem
- Agentic AI and Code Reviews
- Agentic Skill Decay
- Presentation: From AI Agent Demo to Production: Automated Testing and Evaluation
- AI coding creates two kinds of debt. You're only measuring one
- Google's AlphaEvolve Reaches General Availability with Evolutionary Code Optimization as a Service