Why you can't fix the System from within the System + Investment implications
The system isn’t a set of ideals; it’s a set of incentive-compatible control loops designed to conserve itself. That’s why “fixing it from within” almost always becomes absorbed dissent.
I first heard the phrase “You can’t fix the system from within the system“ from a well-known Bitcoiner - Jeff Booth - in the context of politics.
Much of my investment thesis relies on people’s inability to fix the system from within the system, so I’ll go into more detail here.
The system isn’t a set of ideals; it’s a set of incentive-compatible control loops designed to conserve itself. That’s why “fixing it from within” almost always becomes absorbed dissent.
The core lens
incentives > ideals: people and institutions optimize for payoffs, not proclamations.
control > fairness: architectures that preserve discretion beat those that maximize justice.
stability > truth: narratives that keep the machine running outrank facts that would jam it.
revealed preference: watch what budgets, defaults, and penalties do — not what white papers say.
Why “inside” reform gets neutralized
1) Gatekeeping at the permission layer
Credential choke-points (licensing, security clearances, tenure, procurement eligibility) filter out reformers or reshape them early.
Overton window management: staffing, grants, media access go to actors who fit the acceptable range. Deviate and you’re “not serious”.
Revealed preference: careers advance when they reduce variance, not when they increase truth.
2) Selection effects masquerading as merit
Those who remain and rise are those who’ve already internalized the house rules (explicitly or tacitly).
By the time you can “change things”, your payoff matrix is entangled with the institution’s continuity.
Result: reformers become maintenance managers.
3) Co-optation through budget and scope
“We hear you — join the task force.” Your critique is granted a budget and a mandate; metrics are defined by the incumbent.
Goodhart kicks in: you now optimize the metric (deliverables, KPIs, press wins), not the goal. Funding becomes the leash.
Tell: language shifts from verbs (“stop X”) to optics (“raise awareness”, “publish framework”).
4) Perimeter governance beats law
It’s cheaper to enforce via defaults/Acceptable Use Policies (app stores, banks, clouds, ISPs) than statutes.
You “win” a hearing; they update the terms of service. Your reform doesn’t even touch the real switch.
Revealed preference: enforcement migrates to layers you can’t vote on.
5) Complexity by design
Systems accumulate interpretive complexity (exceptions, committees, cross-references) to preserve discretion.
Complexity makes reform illegible to outsiders and makes insiders indispensable.
Consequence: every “fix” becomes another layer, not a deletion.
6) Emergency power as a ratchet
Crises create temporary knobs. Sunsets slide; pilots become defaults.
Appeals to principle are weak against parameterized control that “worked last time”.
Revealed preference: the architecture prizes executable knobs over philosophical limits.
7) Principal–agent geometry
Voters want outcomes; agents want survivable optics and resource control.
When outcomes and optics conflict, optics win — and budgets go up.
Proof: failures produce bigger agencies, not leaner ones. Similarly, scandals often lead to grants for State-Embedded companies.
8) Punishment asymmetry
Dissent is high-friction (career/financing/legal costs); conformity is low-friction (access/promotion).
Over time, the distribution shifts toward compliant narratives, even among the intelligent.
Revealed preference: the machine minimizes the cost of obedience, not the cost of error.
Common “inside” paths to nowhere (recognize them instantly)
Blue-ribbon commissions: delay engines that convert moral energy into PDFs.
Pilotitis: perpetual pilots with no deletion of incumbent processes.
Framework theater: “guidelines” that shift liability to small actors while leaving core choke-points untouched.
Metrics laundering: new dashboards that change measurement, not behavior.
Stakeholder roundtables: consent calibration, not course correction.
What actually shifts outcomes
You don’t “fix” the system by pleading with its ideals; you change its incentive gradients or route around its perimeters.
A) Change defaults
The most powerful reform is a new default (on by default compliance, off by default data sharing, automatic sunset).
If you can’t set the default, you haven’t changed anything. You’ve added a preference pane.
B) Move the budget
Follow the money. If a reform doesn’t reallocate budget/power (from X to Y), it’s decoration.
Force pay-for-performance where failure used to grow headcount.
C) Alter choke points — not slogans
Target the pipes: app-store policies, bank on-ramps, procurement templates, audit requirements.
Write the template everyone will copy. That’s control.
D) Build parallel rails
Create an outside path that’s cheaper/safer/faster for the median actor. Adoption beats argument.
If the parallel rail composes back into legacy (adapters/APIs), incumbents absorb it; design so it displaces instead.
E) Delete, don’t layer
Wins that remove processes/units stick; additions get co-opted.
If your “reform” increases the number of dashboards, you’ve lost.
Reading mass behavior through this lens
People don’t act “against their interests”; they act with the incentives they face right now.
Short-term certainty + low social risk beats long-term benefit + social pain.
The system manufactures those gradients (carrots on supervised rails, friction on sovereign exits) to keep stability > truth.
Translation: education or “awareness” rarely reprograms behavior without environmental rewiring.
Investment edge
Buy the companies that sell the knobs: identity, audit, lineage, revocation, admissibility, and default policy templates — because that’s how control is implemented when consent is scarce.
Avoid vendors that rely on willpower or “principled adoption”.
Trade policy synchronization (same rule showing up across allies) as the leading indicator of revenue consolidation into those platforms.
Bitcoin and Gold are your off-grid hedge for regime breaks.
Personal ops (so you don’t get absorbed)
Design your own defaults: automate the long-term actions, block the short-term dopamine.
Route around permission: pick tools and stacks that don’t strand you when perimeters shift.
Bottom line
You can’t fix a system from inside when its revealed preference is to protect control and stability over fairness and truth. Inside lanes are built to absorb critique, delay change, and ratchet capability. Real shifts come from changing defaults, budgets, and choke points — or from building parallel rails that make the old path economically irrational. Everything else is scripted catharsis.