Incoherent Organizations
Are Large Firms Destined To Be Stuffed With Bullshit Jobs?
Joan Westenberg, whose About page sits somewhere pretty close to the center of the Venn diagram of what this newsletter is all about:
I publish a weekly column on technology, culture, philosophy and what it means to be a human being.
has an interesting take on how AI and the labor market are intersecting. While it still doesn’t bode well for those who want to just have normal corporate employment in the Age of AI, those who are backed into the corner of entrepreneurship may end up with some advantages.
Anyone who has worked at a large, or even medium-sized company has probably had the experience of looking around at the other people who work there and wondering, “what is the point of that person, what do they even do here?” Those with a degree of self-awareness have probably realized that many of their fellow employees are probably looking at them with the same questions.
From the viewpoint of the employee, who knows their job like the back of their hand, these other employees seem superfluous and clueless because they’re existing in some other contextual realm, outside that of said employee. From the viewpoint of the firm overall, though, their continued employment makes complete sense.
In 1937, the British economist Ronald Coase asked a question that seems almost embarrassingly simple: why do firms exist at all? If markets are so efficient at allocating resources, why don’t we just have billions of individuals contracting with each other for every task? Why do we need these hulking organizational structures called companies?
His answer, which eventually won him a Nobel Prize, was transaction costs. It’s expensive to negotiate contracts and coordinate with strangers, to monitor performance and enforce agreements. Firms exist because sometimes it’s cheaper to bring activities inside an organization than to contract for them on the open market. The boundary of the firm, Coase argued, sits wherever the cost of internal coordination equals the cost of external transaction.
Every company has some similar version of the following story:
A company starts small, with the founders doing everything themselves. They make decisions quickly because they understand everything about the business, and the business works.
The company grows and the founders can’t do everything anymore. They hire people and try to transfer their understanding. But understanding doesn’t transfer easily, so they also transfer processes. “This is how we do X. Use this checklist for Y. Follow these steps.”
The processes work, mostly. But the new employees don’t have the context that generated those processes. They don’t know why step three comes before step four, and they don’t know which parts are essential and which parts were arbitrary choices. So when situations arise that the process doesn’t quite cover, they either follow the process rigidly and get suboptimal results, or they improvise and create inconsistency.
More growth, more employees, more processes. The processes start interacting in ways nobody anticipated. The sales process assumes certain things about the product process. The product process assumes certain things about the engineering process. When those assumptions drift out of alignment, you get friction and delays and finger-pointing.
The company responds by adding coordination mechanisms like project managers, alignment meetings, and cross-functional reviews. These help, but they also add overhead, and they create their own drift: the coordination layer develops its own processes, its own assumptions, its own information loss.
Eventually you reach a point where a significant fraction of the organization’s energy goes toward internal coordination rather than actual value creation. A 2022 Microsoft study found that employees in large organizations spend over 50% of their time on internal communication and coordination. Half the payroll, dedicated to getting the organization to agree with itself.
A lot of the AI adoption hype that’s leading to companies laying off employees (even though it’s probably just cover for layoffs they wanted to do anyway) revolves around solving this context probably with technology. But as Westenberg points out, it’s just as likely (or even more likely) to make the problem worse.
More information means the coordination problem gets worse, not better. This seems counterintuitive, because shouldn’t more information make everyone more aligned?
But information isn’t understanding. Understanding = integration, and integration happens in minds. More information means more raw material that each mind has to process differently.
A typical large organization’s knowledge base is spilling over with strategy documents from last year and the year before, project postmortems from dozens of initiatives, customer research reports, competitive analyses, technical specifications, meeting notes, email threads, Slack channels, and wiki pages.
Somewhere in there (the elusive somewhere...) is everything you need to know to make a good decision.
But nobody has synthesized it all, and nobody has integrated it into a coherent model. Each person reads a fragment, interprets it through their own context, and forms their own understanding. When they discuss decisions with colleagues, they’re not comparing the same mental models but rather different interpretations of different subsets of the available information.
This is context fragmentation. People don’t disagree on facts; they’re operating from different maps of the same territory. And because the maps are implicit, inside people’s heads, nobody realizes they’re not looking at the same thing.
The proliferation of AI tools in large organizations means that now each employee has their own AI assistant, trained on whatever context they happen to feed it, producing outputs that reflect their particular understanding of the situation. The AI amplifies individual perspectives rather than creating shared ones.
I’ve always thought of companies as facing a constant tension between centralization and federalization. The centralization impulse comes from wanting to keep coherence tightly controlled, which should lead to better decision making. The federalization impulse comes from realizing that when you stay centralized to make better decisions you end up not making enough of them — too many things get left undone. The result of this tension manifests as constantly shifting organizational designs, an ebb and flow of layers of management, and a never-ending stream of new technology solutions that promise to solve the problem. AI, as currently being utilized, is the latest one of these.
All this means that a solo entrepreneur has certain advantages.
When you’re operating alone, you have one context, one understanding, one model of your business and your market and your customers and your strategy. That model lives in your head, and it’s coherent because there’s only one mind maintaining it.
If // when you use AI tools, you’re feeding them from that single source of truth. The AI doesn’t have its own understanding that might drift from yours, and it operates within the context you provide. If you give it good context, it executes within that context. If your understanding is coherent, the AI’s outputs will be coherent.
This is the inversion of the traditional organization’s problem. In a large organization, you have many minds with their own contexts, trying to coordinate through AI tools that amplify their differences. As a solo operator, you have one mind with one context, using AI tools to execute within that coherent frame.
The AI handles the execution at scale while you maintain the coherence. This division of labor plays to the strengths of each party: humans are good at integration and judgment, while AI is good at execution and volume. The solo operator with AI gets the benefits of scale without the costs of coordination.
And because of this, smaller operations can punch above their weight and be competitive.
Scale used to be the moat. You built a big organization with lots of resources, and the sheer weight of your operation protected you from smaller competitors. Transaction costs made it hard for anyone to replicate what you’d built.
But transaction costs are collapsing. The activities that used to require organizations can increasingly be performed by individuals with the right tools.
The Coasean logic that justified large firms is weakening.
If there is a new moat (and I’ll admit, that’s a big “if”) it probably looks like coherence; the ability to operate as one mind, one understanding, one model, even as you execute at scale. Large organizations can’t have this because they’re composed of many minds. Solo operators can have it by default, if they’re deliberate about maintaining it.
The solo operator who builds a coherent system and lets AI execute within it has an advantage that doesn’t need venture capital, doesn’t need hiring, and doesn’t ask for the whole apparatus of organizational scaling.
Scale breaks coherence. Coherence is the moat.
As a reluctant entrepreneur myself, it’s comforting to know that I’m actually better off than I was when I had a job. But something that strikes me is that the article’s description of how large organizations function describes it as things currently are, not as they possibly could be.
Yes, the coordination and differing context problems are real, as anyone who has had a job can attest, but the same tools that allow the solo entrepreneur to execute at scale also exist for the large firm to do so. If the large firm could just crack the coherence problem, they wouldn’t need to be large (in terms of bullshit employees) anymore.
I’m not really sure what an Obsidian vault is, but I know enough to know that there are existing technologies and workflows that purport to solve your coherence problem. There is nothing to prevent an organization from indulging in their centralizing organizational impulse by enabling AI to run from a centralized context rather than the many different distributed contexts Westenberg describes in her article.
This is what AI adoption means. It’s not really that difficult if you decide to do it (AskUserQuestion at scale). Once a few large companies figure this out the rest will as well. And when that happens, bye-bye jobs.
But at least everyone will be able to be a company unto themselves.
News Scan
ECB Holds at 2.15% but Signals April Hike; Nagel, Makhlouf, and Villeroy Line Up Behind Tightening
Source: Bloomberg / CNBC / ECB
Date/Time: March 19-20, 2026
The ECB held rates at its March 19 meeting but the tone was unmistakably hawkish — the most dramatic shift since the war began. President Lagarde warned the outlook is “significantly more uncertain” and raised the 2026 inflation forecast to 2.6%, while publishing a severe scenario showing euro-zone inflation peaking at 6.3% in Q1 2027. Within 24 hours, three Governing Council members publicly opened the door to an April hike. Bundesbank President Nagel said the ECB “will need to consider” tightening if the inflation outlook deteriorates further. Ireland’s Makhlouf said an April hike is “possible if data signal need.” France’s Villeroy declared the ECB is “totally determined” to meet its inflation target. Traders responded by fully pricing three quarter-point hikes this year — a complete reversal from the easing expectations that prevailed before the war. Lagarde separately urged governments to keep energy aid “temporary, targeted, and tailored,” explicitly warning against a repeat of the blanket fiscal response seen during the 2022 Ukraine crisis. The ECB has pivoted from a cutting cycle to openly discussing hikes in under three weeks — the fastest reaction function shift in its history.
Bloomberg
BOE Holds at 3.75% Unanimously, Dovish Wing Flips Hawkish; JP Morgan Expects April Hike
Source: Bank of England / Bloomberg / LBC
Date/Time: March 19, 2026
The Bank of England voted unanimously to hold rates at 3.75% — its first unanimous decision since December 2021, and a striking signal of committee alignment. The real story is the shift within the MPC’s dovish wing. Catherine Mann, who had been pushing for cuts, said her view had shifted “away from considering a rate cut and towards a longer hold, or even a hike.” Even Swati Dhingra — the committee’s most persistent dove — acknowledged rates may need to rise if oil and gas supply disruption persists. The BOE now expects inflation to exceed 3% for most of the year, potentially reaching 3.5% in Q3. JP Morgan moved its BOE call to hikes in April and July. Before the war, markets had been pricing the BOE’s next move as a cut. UK gilt yields hit their highest level since 2008, reflecting the market’s rapid repricing of the rate path.
Bank of England
UK Posts £14.3B Budget Deficit — Nearly Double the OBR Forecast — on Record Debt Interest
Source: Bloomberg / ONS
Date/Time: March 20, 2026
The UK government borrowed £14.3 billion in February, far exceeding the £8.8 billion economists had forecast and nearly double what the Office for Budget Responsibility projected in November. The overshoot was driven by record-high debt interest payments for the month of February, even before the full impact of the Middle East war hit gilt markets. This data point directly constrains the BOE’s policy space: the government’s fiscal position is deteriorating precisely when the central bank may need to raise rates, which would further increase debt servicing costs in a country with a high proportion of inflation-linked gilts. The feedback loop — higher energy prices push up inflation expectations, which raise gilt yields, which increase debt costs, which widen the deficit — is the UK’s core macro vulnerability in this crisis.
Bloomberg
China Government Spending Hits Fastest Pace Since 2022 as Beijing Front-Loads Stimulus
Source: Bloomberg / SCMP / CNBC
Date/Time: March 19, 2026
China’s broad measure of public expenditure climbed 6% year-on-year in the January-February period — the fastest start to any year since 2022 — as authorities moved to fortify the economy against rising external uncertainty. Retail sales rose 2.8% YoY, beating the 2.5% consensus and accelerating sharply from December’s 0.9%, boosted by extended Lunar New Year spending and pro-consumption policies. The fiscal acceleration is the most important variable in determining whether China can escape its deflationary trap. Beijing is front-loading stimulus to build a buffer before the Iran conflict’s second-round effects hit Chinese manufacturing and exports. However, the conflict may complicate the playbook: analysts warn it could flip China’s deflation problem into “bad inflation” — cost-push price increases without corresponding demand — which would narrow the PBOC’s room to cut rates further and undermine the very stimulus effort the spending data reflects.
Bloomberg
QatarEnergy CEO Confirms LNG Damage Will Take 3-5 Years to Repair — 17% of Capacity Offline
Source: CNBC / Bloomberg
Date/Time: March 19, 2026
QatarEnergy CEO Saad al-Kaabi confirmed that Iran’s missile strikes on Qatar’s Ras Laffan complex have knocked out approximately 17% of the country’s LNG export capacity, and repairs will take three to five years. This transforms what markets initially treated as a wartime disruption into a structural, multi-year supply deficit. Qatar provides roughly 20% of global LNG trade, and the damaged capacity cannot be quickly replaced — there are no idle LNG trains anywhere in the world operating at the scale needed. European gas markets, already pricing in longer disruption, face the prospect of sustained elevated prices well beyond any ceasefire. The TTF benchmark remains sharply above pre-war levels. For Asia, the hit to contracted Qatari supply forces buyers into an already-tight spot market, competing directly with European buyers and driving up LNG prices globally.
CNBC
IEA Declares Supply Loss Exceeds 1973 Oil Shock, Calls for Emergency Demand Measures
Source: Bloomberg / IEA / OilPrice.com
Date/Time: March 20, 2026
The International Energy Agency stated that the current supply disruption is now larger than the 1973 oil shock that led to the IEA’s creation — and larger than any disruption since. In its March Oil Market Report, the IEA cut its 2026 global demand growth forecast by 210,000 barrels per day and called on governments and consumers to implement emergency demand-reduction measures: working from home, slower highway speeds, greater use of public transport, reduced air travel, and expanded electric cooking. The framing is significant — the IEA is the institutional voice of energy-consuming nations, and its explicit comparison to 1973 signals that the agency views this as a crisis of historic proportions requiring coordinated policy responses, not just market adjustment. The demand-side recommendations echo the IEA’s 10-point plan from 2022 but with greater urgency.
IEA
Oil and Fuel Cargo Premiums Smash Records as Refiners Scramble to Replace Middle East Crude
Source: Bloomberg / Reuters / BOE Report
Date/Time: March 19-20, 2026
The supply stress from the Hormuz disruption is now fully visible in physical crude markets, with premiums reaching levels never seen before. Norwegian Johan Sverdrup — a medium-sour crude that can partially substitute for Gulf barrels — was bid at a record $11.30 per barrel premium to Brent, implying a cash price around $124. Russian Urals crude delivered to India has reversed from a $13 discount to a $4-5 premium over Brent — the first time Urals has traded at a premium in modern history. In the Middle East, premiums of up to $40 per barrel above benchmark are being quoted. US Mars Sour hit $107.53, its highest since 2008. Refiners globally are holding off on purchases because margins are being destroyed by these premiums, leading to reduced crude processing and amplifying the shortage of refined products downstream. The premium data confirms that strategic petroleum reserve releases and non-OPEC supply increases are insufficient to offset the Hormuz disruption.
Reuters via BOE Report
Russia Cuts Key Rate to 15%, Benefiting From the Very Oil Shock Pressuring Everyone Else
Source: Bloomberg / Moscow Times
Date/Time: March 20, 2026
The Bank of Russia cut its benchmark rate by 50bp to 15.0%, continuing a cautious easing cycle that began last October from the peak of 21%. Annual inflation has slowed to 5.9% as of mid-March, giving policymakers room to move despite the global uncertainty. The irony is stark: while the ECB and BOE are pivoting toward hikes because of oil-driven inflation, Russia — as a major oil exporter — is cutting because the same shock is boosting its revenues, strengthening its fiscal position, and allowing the economy to cool from overheating. The US has also partially lifted sanctions on Russian oil to increase global supply, further improving Moscow’s terms of trade. Russia is one of the clearest beneficiaries of the conflict’s commodity price impact, creating a widening policy divergence with Western central banks.
Bloomberg
Brazil Finance Minister Haddad Resigns to Run for São Paulo Governor; Durigan to Succeed
Source: Bloomberg / Reuters / US News
Date/Time: March 19, 2026
Fernando Haddad formally stepped down as Brazil’s finance minister to run for governor of São Paulo, the country’s most populous state. His deputy Dario Durigan is expected to take over. Haddad’s departure comes at a delicate moment — Brazil just began its easing cycle with a 25bp cut to 14.75% on March 18, and the country’s fiscal trajectory remains a key concern for markets. Haddad had led the overhaul of Brazil’s consumption tax system, long seen as a structural drag on growth, but faced persistent criticism over rapidly rising public debt. The transition to Durigan signals policy continuity, but the loss of Haddad’s political capital within the Lula government — and the distraction of an election campaign — adds uncertainty to Brazil’s fiscal consolidation narrative at exactly the moment when the oil shock is complicating the macro outlook.
Bloomberg
Disclaimer: The information provided on this blog is for informational purposes only and should not be considered investment, financial, or other professional advice. Nothing on this site constitutes a recommendation or solicitation to buy or sell any securities. You should consult with a qualified financial advisor before making any investment decisions. Investing involves risks, including loss of principal.