You Are More Powerful Than You Know

Sauna strife

In mid-April 2026 I ordered a Luxo Living outdoor Finnish sauna for delivery to my Brisbane home. Total: $3,887.10. The website quoted delivery by 23 May. My order carried an explicit written instruction: if you cannot deliver before 15 May, do not deliver before 7 August, because I will be overseas for two months from late May. Standard stuff. The kind of thing you write once and expect a competent supplier to handle.

Luxo could not meet the window. Stock was delayed. Their response was not "we apologise for the delay and will deliver on your return as instructed." Their response was "we will charge you $5 per day storage while we hold the goods we failed to deliver on time." Roughly $350 in storage fees for their failure to perform, with the implicit threat that non-payment would mean forfeiture or further escalation.

This is a small dispute. $350 is not a life-changing sum. The sauna itself is not a life-changing purchase. The reason it matters — the reason it's worth starting an essay with — is that the entire interaction is a compressed instance of how scaled organisations extract compliance from individuals who have been trained to believe they are powerless. The machinery is the same whether the stakes are $350 or $350,000. The difference is only in whether the individual notices the machinery is operating on them.

Most people pay the $350. Not because they agree with the charge. Because disputing it feels harder than paying it, because the organisation's policies are presented as if they were law, because the power asymmetry is performed so convincingly that the individual accepts it as real. The extraction works by persuading you that your only choices are comply or give up the goods.

The purpose of this essay is to name the machinery, give the generic playbook for refusing it, and then work through what "refusing it" actually is at the cognitive level — because the playbook is the easy part, and the hard part is noticing when the playbook needs to run.

The generic playbook

When a scaled organisation applies pressure to extract something they are not entitled to, the moves that reliably work to stop them doing this are not secret. They are not complicated. They are simply not taught, because teaching them would reduce the extractive capacity of the system that benefits from most people not knowing them. Here are the moves:

Find a bigger bully. Every consumer interaction in Australia sits under the Australian Consumer Law. Every state has a Fair Trading regulator. Every industry has an ombudsman or equivalent. The ACCC exists. The bigger bully is the regulator whose threat the corporation actually respects. Naming the bigger bully in correspondence — "I will file a complaint with Queensland Fair Trading under sections 60 and 62 of the ACL if this is not resolved by [date]" — changes the calculation on their side from "extract $350 from this customer" to "defend against a regulatory complaint that could cost more than $350 to respond to even if we win." Most disputes end at this step because the economics of corporate compliance favour settlement over regulatory engagement.

Keep great records. Every email, every website screenshot, every order confirmation, every shipping notification, every phone call logged with time and name. Scaled organisations rely on information asymmetry — they have records and you do not, so their version of events becomes the default. Neutralise this by keeping records at least as good as theirs. A PDF archive of the relevant website pages at the moment of purchase is worth more than any later argument about what the website did or didn't say.

Make an excellent legal argument with LLM help. This is new and it matters enormously. Five years ago, an individual disputing a $350 storage fee could not afford the legal specificity that makes corporate lawyers treat a complaint seriously. Today, you can draft a Fair Trading complaint that cites specific sections of the ACL, identifies the corporate entity's specific breach, quantifies loss correctly, and anticipates the standard defences, in minutes, for close to free. The corporate side no longer has the monopoly on legal specificity that used to protect them. Use it.

Create a BATNA. Best Alternative to Negotiated Agreement. If the dispute cannot be resolved, what will you do instead? For the sauna, my BATNA was clear before I wrote a single dispute email: equivalent AS/NZS-certified outdoor Finnish saunas from alternative Australian suppliers cost approximately $7,000. The loss if Luxo repudiates the contract is not $3,887 — it's the $3,100 differential between what I paid and what the substitute costs. A second BATNA is also available and worth knowing: I could pay someone to be home and accept delivery on my behalf while I am overseas, perhaps two hours at $35 per hour, roughly $70. That's a fraction of the $350 storage fee Luxo is demanding, and it is an option I can unilaterally execute without Luxo's cooperation. Having two BATNAs — one for full repudiation, one for partial cooperation — means you cannot be bluffed by "well, you can always just cancel" or by "we have no choice but to charge storage." Cancellation is not costless to them once the differential is established and documented, and storage is not the only solution to the delivery problem.

Refute all corporate attempts to claim their policies are the defining mechanism. This is the move they rely on most heavily and it deserves careful attention. The corporate script runs: "Our terms and conditions say storage fees apply after 14 days." Or: "By checking the box at checkout you agreed to our policies." Or — most commonly — "When we sent your order confirmation, it included updated terms that cover this situation."

All of these are attempts to retrofit contract terms that either were not in the original agreement or are unenforceable as written.

The contract between you and the supplier consists of the offer as presented at the point of sale, the acceptance by payment, and the terms that were genuinely incorporated by reference at that moment. A tiny checkbox linking to a 40-page T&C document buried three clicks deep is not reliable incorporation of every clause in that document — courts routinely find unusual or onerous clauses unenforceable when they were not brought to the customer's attention. More importantly, terms that the corporation sends back in their order acceptance or later correspondence that were not in the original offer are not binding on you. A contract is formed at the point of acceptance. Adding terms afterwards is an attempt at unilateral variation, which is not how contract law works. "Our invoice says..." is a sentence that reveals the attempt rather than establishing the term.

And even where terms were properly incorporated, sections 23–28 of the ACL render unfair contract terms void in consumer contracts. A clause that imposes storage fees on the consumer for the supplier's own delay is a paradigm example of an unfair term — it causes significant imbalance, it is not reasonably necessary to protect the supplier's legitimate interests, and it would cause detriment to the consumer if applied. Internal policy is not contract. Contract is not statute. Statute overrides contract where the contract is inconsistent with statutory consumer guarantees. The corporation will present its policy as if it were the governing authority. It is not. The governing authority is the law. This is where the LLM comes into its own. You can tell it what happened, share all the written documents, screenshot the website, and it knows the law. What used to require a lawyer on retainer now requires a conversation and some attached files.

Refute all corporate attempts to avoid responsibility for their part of the conflict. "Stock delays are outside our control." "Our supplier let us down." "Shipping delays are a known industry issue." All of these are irrelevant to you, the consumer. Your contract is with the supplier. Their contract with their suppliers is their problem. The cause of the delay does not change who owes the remedy. When they try to diffuse accountability by pointing upstream, bring it back: "I am not a party to your supplier arrangements. My contract is with you. The breach is yours."

This playbook resolves most disputes of this type. It takes some commitment to execute, but the payoff can be worth it. It requires zero specialised training that an LLM cannot supply, and it flips the power asymmetry almost entirely. The corporation was relying on you not running this playbook. When you run it, the dispute becomes expensive for them and cheap for you, and they settle.

But why don't most people run the playbook?

This is where the essay gets interesting, and where it stops being about saunas and storage fee scams.

Most people, most of the time, in most of the domains of their lives, do not run the playbook even when they clearly know the playbook exists. They know they could dispute the charge. They know the regulator exists. They know they are being extracted from. They pay anyway. Not because they can't — because something upstream of the ability-to-act never fires.

The missing factor is not capacity. It is not courage. It is not even knowledge, strictly speaking. It is the specific moment of noticing: this is a situation that requires me to think carefully about what is being done to me. That moment of noticing is the seed of agency, and it is what the rest of this essay is about.

One thing worth stating clearly before we get to mechanism. We all need to run the playbook. Not only because it serves us individually, but because extractive business models only stop being profitable when enough of the target population refuses to comply. The storage fee works on the assumption that 90% will pay. If that number drops to 50%, the model becomes unviable and the organisation restructures. If it drops to 20%, the fee disappears. Every individual who runs the playbook tilts the economics fractionally against extraction for everyone else who comes after them. Be the hero. Don't take one for the team — make THEM take one for our team. The compliance default is the team taking a shot to the wallet; refusal is the team refusing to. The moral frame is inverted from how it's usually presented. Disputing the $350 is not selfish pedantry. It is the small civic act of not paying a rake that was never legitimately imposed.

Where agency actually comes from

The standard folk-model of agency treats it as a single capacity you either have or don't. Strong-willed people have it, weak-willed people lack it. Free will enthusiasts say everyone has it; determinists say nobody does. Either way it's a scalar — more or less of a single thing.

This model is probably wrong, and the way it's probably wrong is structurally important. Agency is not a scalar. It is plausibly a product of three separable factors, each with different properties, each with different development paths, and each capable of being the single broken piece that makes the whole thing collapse.

Factor one: awareness that meta-awareness is needed in this situation. This is the trigger. It's the moment you notice that automatic processing is about to run and should be interrupted. For the Luxo dispute, this is the moment "I should just pay and move on" gets interrupted by "wait — should I be paying this at all?" For most people, in most domains, that interrupt never fires. They just pay. This factor is trainable, domain-specific, and sits in the skill category. It is also the thinnest layer of the three — which is interesting, because it's also the one most people never work on.

Factor two: meta-awareness capability. Once the trigger fires, how well can you actually examine what's happening? How clearly can you represent the situation abstractly, hold multiple hypotheticals in mind, notice your own reasoning patterns, detect where you're being manipulated? This is closer to a trait. It has a baseline component that is partly heritable and partly developed early. It possibly can be sharpened through training (cognitive psychology education, philosophical practice, therapy), but it is not clear if these things simply improve the trigger for meta-awareness rather than the capability itself. The observed improvement from training might be entirely attributable to better triggers firing more often on a fixed underlying capability, rather than the capability expanding. The distinction matters because it affects what we can honestly promise about developmental interventions.

Factor three: ability to act on what meta-awareness reveals. This is the override capacity. Seeing clearly is not the same as being able to move. Addicts demonstrate the separability of this factor with painful clarity — they see what they're doing, they know what they should do, and the action capacity is nevertheless captured by the addiction mechanism. The ability to act is partly skill (strengthened through successful use) and partly subject to domain-specific capture (addiction, trained reflexes under stress, learned helplessness in particular relationships).

The product model

Let's consider that agency in any given domain could be the product of these three factors, not their sum:

Agency = Trigger-awareness × MA capability × Ability to act

The multiplicative structure matters because of the zero property. If any single factor is zero, agency in that domain is zero. No awareness that you should be checking → you don't check. Unable to represent the situation clearly even when checking → checking produces nothing useful. Unable to act on what you see → seeing produces only suffering.

Most failures of individual agency in scaled-organisation extractions are failures of the first factor. The trigger never fires. People have the cognitive capability to see the extraction clearly if they looked, and the action capacity to refuse it if they saw it, but the initial "wait — should I examine this?" moment doesn't occur. The compliance is automatic. The organisation's design relies on this precise failure — they are not defeating your analysis, they are ensuring analysis never begins.

The multiplicative structure also means any non-zero increase in any factor increases the product. You do not have to be radically more agentic across the board to make a meaningful difference in a specific domain. You only have to install one trigger in one domain. The product coming from the combination of all three things moves. You are more agentic there, permanently, once the trigger is in place.

The actionable bit

The good news is that the layer most people are missing — trigger-awareness — is also the cheapest layer to train. It does not require years of cognitive psychology education. It does not require psychotherapy. It requires one thing: deliberate priming of your attentional system to notice specific situations as situations that warrant examination.

The mechanism is the Reticular Activating System. The RAS is the filter that decides what reaches your conscious processing. It is primed by deliberate attention. If you decide, explicitly, that "interactions with scaled organisations that involve demands for money" is a category that warrants examination, the RAS begins tagging those interactions in real time. The tag is the trigger. Once the tag fires, meta-awareness engages automatically for people with adequate capability in that factor.

This is why implementation intentions work so well in the psychological literature — effect sizes around d=0.65 across hundreds of studies, which is large by behavioural-intervention standards. The pattern is: "When situation X occurs, I will do Y." That sentence, repeated and rehearsed, installs a domain-specific RAS trigger. You are not trying to be more agentic in general. You are installing a specific ignition condition for a specific category of situation.

For the Luxo-class of situation, the implementation intention is something like: "When a scaled organisation demands payment I did not clearly agree to, I will pause and ask what options I have other than just paying in this situation before I respond." That's it. That's the whole training. Repeat it until it fires automatically when the situation arises.

One caveat worth flagging: installing triggers in domains where your ability-to-act is captured by something else — addiction, genuine coercion, power differentials you truly cannot move against — will produce suffering rather than agency. Sadly, for most of us there is no ombudsman who can take on the ATO, because our taxes are themselves extractive. Trying to is probably just going to cause you suffering. Better to leave the country and pay more reasonable taxes elsewhere — but that is a different third thought. The training works when the other two factors are adequate and the trigger is the missing piece. For most of the scaled-organisation extraction problem, the other two factors are in fact adequate for most people. The missing piece really is the trigger.

The conclusion that matters

The scaled organisations that shape most of modern economic life rely on a specific assumption about individuals: that the default will be compliance because the trigger to examine will not fire. They are usually right about this, which is why they have built extraction into so many of their standard operating procedures. The storage fee on the undelivered sauna is not a bug in Luxo's process. It is built into their business model on purpose to extract more profit on some transactions. Most people pay it. The feature works.

The response is not to become a more agentic person in general. That framing is incoherent under the product model — agency is not a scalar you can globally inflate. The response is to audit your own domains, identify the categories where triggers are currently missing, and install them deliberately one at a time through RAS priming. This is small, repeatable, and available to anyone with enough meta-awareness capability and ability to act to make the other two factors non-zero — which, for the vast majority of scaled-organisation extractions, is almost everyone.

Everyone can increase their own agency. Not by becoming more agentic in some global way. By training, domain by domain, the specific habit of asking: do I need to be meta-aware in this situation? That question, asked in the right moments, is the whole move. Everything the playbook does downstream is only possible if the question gets asked. Once it is asked, the rest follows naturally.

The saunas will be delivered when the contract says they will. The storage fees will not be paid. The organisations will adjust, slowly, to populations that notice. The first step is the question, and the question is free. Importantly, this was not possible before — the corporates could afford lawyers and consumers could not. Now you don't need a lawyer. You just need awareness and an LLM.