We are living in a Moonage Daydream
In the age of autonomous AI, corporate hierarchies are mathematically unraveling as solo builders outcompete bureaucratic firms.
Lately, I find myself unable to stop listening to Moonage Daydream.
It has been playing on repeat, over and over, until the track stops being a song and begins to feel like an atmosphere.
David Bowie was never merely contemporary; through his sound and his lyrics, he operated as if he were broadcasting from a future that hadn't yet bothered to negotiate with the present. His orbit around Andy Warhol, his immersion in the The Factory, and his radical synthesis of pop art didn’t just reflect modern culture—it forcibly bent its trajectory.
Bowie understood before almost anyone that when the medium mutates, the human silhouette must reinvent itself to survive inside it.
Now, it is technology’s turn to mutate. Everything is shifting beneath our feet. The world as we have inherited it is quietly disintegrating, and most people haven't even registered the tremor.
Before this moment, there was another world. For nearly a century, monolithic corporations made coherent economic sense—they were necessary, defensive shelters against crushing transaction costs. But in the age of artificial intelligence, that entire architecture is losing its reason to exist.
While autonomous models act as an exponential multiplier on the individual producer, internal corporate bureaucracy continues to compound at a super-linear drag. At that exact intersection, cooperative game theory delivers an inescapable verdict: the Core empties. A wage vector that simultaneously preserves the solvency of the firm and satisfies the outside options of its best talent becomes mathematically impossible.
The layoffs we read about every morning are not a cyclical downturn. They are not a seasonal corporate belt-tightening. They are the thermodynamic dissolution of an institution whose math no longer holds.
In this essay, we will trace that rupture—from a distorted 1972 guitar solo to the game-theoretic unbundling of the firm, and toward the brutal, uncapped frontier awaiting those who choose to produce.
I. The Death of the Manor: From Coase to Core Emptiness
In 1937, Ronald Coase asked why corporations exist if markets are efficient.
His answer: transaction costs.
Searching, contracting, and enforcing on open markets was expensive, so firms economized on friction through hierarchical command.
Yet Coase’s equation contained an internal trap: bureaucracy (τ) scales super-linearly with headcount.
Model the firm as a Cooperative Transferable Utility (TU) Game (N, v). Pre-AI, coordination friction followed Metcalfe’s and Brooks’ laws: τ(S) = c |S|β with β ≈ 2, driven by the n(n−1)/2 communication channels. Net corporate surplus was:
Because an isolated worker had negligible standalone output (v({i}) ≈ 0), the game was balanced. By the Bondareva-Shapley Theorem (1963, 1967), the Core was non-empty (Core ≠ ∅). A stable wage vector existed satisfying budget balance and individual rationality.
Now enter autonomous intelligence: an exogenous cognitive multiplier A ∈ [1, ∞) compounding exponentially (A(t) = A₀ eμt).
Crucially, A does not reward corporate scale. For a singleton (|S| = 1), internal friction is zero. An elite builder equipped with autonomous agents produces:
Meanwhile, the corporation absorbs A through compliance filters and committee sludge (η < 1), while quadratic drag c nβ persists:
where F(N) = γ ∑ fᵢ (γ ≥ 1).
By Bondareva-Shapley, Core stability requires:
The moment A crosses the critical tipping point:
the Core becomes completely empty (Core = ∅).
An empty Core means no compensation scheme can satisfy both corporate solvency and individual rationality. A builder’s marginal contribution (Shapley value ϕᵢ ~ O(A)) outstrips compressed corporate pay wᵢ = w̄ + θ(ϕᵢ − w̄) (Frank 1984). Elite talent defects (Sexit = {i : A fᵢ > wᵢ}); corporate output collapses while fixed bureaucracy cannot shrink; the Grand Coalition shatters.
Layoffs are not an accounting cycle. They are the thermodynamic dissolution of an empty Core.
II. The Rhymes of History: Shocks That Birthed Producers
Institutions never voluntarily surrender their moats. Every major leap in autonomous production has been triggered by an external shock that destroyed an institutional shield and forced labor into the market.
The economic turmoil of the 1970s accelerated a major transformation in how American businesses were organized. Stagflation, characterized by high inflation and weak economic growth, was followed by aggressive monetary tightening under Federal Reserve Chairman Paul Volcker, with interest rates reaching approximately 20% in the early 1980s. As borrowing became more expensive and economic conditions deteriorated, large conglomerates faced increasing pressure to restructure, sell underperforming divisions, and reduce their workforces.
This period also exposed fundamental weaknesses in how large corporations allocated capital internally. As Michael Jensen (1993) explained, corporate executives often continued directing resources toward inefficient or declining business units rather than investing in more productive opportunities. The restructuring that followed resulted in substantial layoffs and encouraged skilled employees, including engineers and managers, to pursue opportunities outside traditional corporate structures.
At the same time, an important regulatory change made it easier for these emerging businesses to access funding. In 1979, a reinterpretation of the Employee Retirement Income Security Act (ERISA) clarified that pension funds could invest in venture capital as part of a prudently diversified portfolio. This helped channel substantial institutional capital into the venture capital industry (Gompers & Lerner, 1999).
Together, these developments contributed to a shift in the organization of innovation. Instead of relying exclusively on large corporations to finance and develop new technologies internally, engineers and entrepreneurs increasingly had the opportunity to establish independent companies backed by venture capital. This changing economic and financial environment helped support the growth of Silicon Valley's hardware and software startup ecosystem.
III. The Arithmetic of the Mutation
Empirical data confirms the rupture.
Tyna Eloundou et al. (Science, 2024) mapped language models across 1,016 occupations. Unlike industrial automation, which substituted manual labor, LLM exposure is positively correlated with wages and education. Eighty percent of U.S. workers have at least 10% of tasks exposed; 19% have over 50% exposed. The corporate castle was built to house symbolic manipulators; symbolic manipulation is now priced in fractions of a cent.
Daron Acemoglu (2024) estimated a modest 10-year TFP gain of 0.66%, arguing that only 4.6% of tasks can be profitably automated. But as Philippe Aghion et al. (2024) showed, Acemoglu’s model measures only the intensive margin—performing old tasks cheaper. It misses the extensive margin and the compounding of the idea production function:
As hypothesis testing and software synthesis drop toward zero cost, firm creation accelerates.
*U.S. Census Bureau data shows High-Propensity Business Applications sustained at a record 1.8 million annually post-2022. While Alphabet generates ~$1.6M in revenue per employee, Midjourney generates ~$7.5M+ ARR per head with forty people. The minimum efficient scale of the enterprise has collapsed from ten thousand to five.
The Brutal Selection
Now we must address what techno-optimist brochures omit.
A corporate salary was never pure compensation; it was a variance-reduction subsidy. The firm absorbed market volatility and handed you a smoothed median. When you leave the manor, that subsidy vanishes. You are cast onto a heavy-tailed, Pareto distribution.
Entrepreneurship, however, remains far from an easy path. Most businesses struggle to survive over the long term, and even successful founders often earn less than they might in traditional employment. As technology makes building products increasingly accessible, the challenge is shifting from creating something to finding customers, earning their trust, and building a sustainable business. Starting a company may be easier than ever, but succeeding remains exceptionally difficult.
The market does not promise you will win; it only strips away the roof that kept you from the storm.
Far Out in a Moonage Daydream
At the end of Moonage Daydream, beneath fading feedback, Bowie laughs quietly in the studio.
It is the laughter of someone who looked at "The Church of Man"—the corporate cathedral of thirty-year careers, performance reviews, and managerial titles—and realized its pillars were hollow.
"The church of man, love, is such a holy place to be
Make me freak out, far out, in a moonage daydream"
The Core is empty. Layoffs will continue because maintaining the Grand Coalition is mathematically irrational.
The sovereign producer knows the bargain. They do not fear, because they understood before scaling the rock that risk is the price of uncapped upside.
The stars look very different today.
And lastly, I want to finish this essay with my favorite lyrics:
Keep your 'lectric eye on me, babe
Put your ray gun to my head
Sources & Notes
Cooperative Games & Firm Theory: Ronald Coase (1937); Oliver Williamson (1975). Balanced sets and the Core: Lloyd Shapley (1967); Olga Bondareva (1963). Wage compression: Robert Frank (1984).
Historical Macro Shocks: Black Death: M.M. Postan (1973); Robert Brenner (1976); S.R. Epstein (2000). Enclosures: Karl Polanyi (1944); Robert Allen (1992); Franklin Mendels (1972). De-conglomeration: Michael Jensen (1993); Paul Gompers & Josh Lerner (1999). Dynamo paradox: Paul A. David (1990).
Empirical AI & Entrepreneurship: Labor exposure: Tyna Eloundou et al., Science (2024). AI macro: Daron Acemoglu (2024); Philippe Aghion et al. (2024). Entrepreneurial earnings penalty: Barton Hamilton, JPE (2000). Founder returns: Robert Hall & Susan Woodward, AER (2010). Attention economics: Herbert Simon (1971).
Culture: David Bowie, "Moonage Daydream," Ziggy Stardust (1972). Studio documentation: Ken Scott & Bobby Owsinski (2012).