The Loop, the Abyss, and the Denial Reflex

Semiconductors, AI, and the Arithmetic of Financial Irrationality

There are moments when the market appears to look directly into the abyss – and then immediately pretends it is not there. The recent sequence of record‑scale market cap swings in Apple and Nvidia is one of those moments

What these events reveal is not merely volatility, but a deeper irrationality: a system that prices trillions of dollars of “future value” while actively suppressing any serious reckoning with the physical, social, and geopolitical constraints that my previous article, The Loop and the Black Hole, sought to map.

I. Trillions in a Day: The Numbers That Don’t Add Up

In the last few years, single‑day moves in the market capitalization of mega‑cap tech have reached scales that would previously have been considered almost science fiction. Apple has registered one‑day losses on the order of 180 billion dollars, followed by multi‑day routs approaching several hundred billions, only to recover and climb towards, and then above, a 5 trillion dollar valuation. Nvidia’s market cap, in turn, has surged past 3 and then 4 trillion, with commentary now entertaining the possibility – and temporarily the reality – of a 5 trillion valuation centred almost entirely on AI‑related demand for GPUs. https://finance.yahoo.com/video/apples-market-cap-tops-5t-194827598.html  

These numbers are not marginal. A single day’s swing in a single stock can now exceed the total GDP of many mid‑sized economies, or the entire market capitalization of the vast majority of companies in the S&P 500. In The Loop and the Black Hole, I argued that the semiconductor sector as a whole trades at around 41.7 times its gross value added (GVA), with an implicit price‑to‑earnings ratio near 200, and a market‑cap‑to‑wage‑bill ratio of ~58: for each dollar paid in wages, the market attributes 58 dollars of equity value.

To make those multiples coherent, we showed, the industry would need to multiply its profits by roughly 38 times in twenty years and anchor an ecosystem generating on the order of 28 trillion dollars of GVA – about 15% of projected world GDP – in a complex “semiconductors + AI + automation” that displaces and absorbs a share of economic activity comparable to the entire global manufacturing sector today.

When one trillion of Apple’s valuation evaporates and reappears over a period of weeks, or when Nvidia adds hundreds of billions in a single session, what is being repriced is not just a company’s future cashflow. It is the probability assigned to that extraordinary reconfiguration of the world economy – and yet the political, physical and social implications of that reconfiguration remain almost entirely absent from mainstream financial analysis.

II. Market Cap as Wages: Volatility as a Human Problem

One of the central claims of The Loop and the Black Hole was that, in high‑multiple sectors, market capitalization functions as a wage instrument rather than a passive measure of value. In semiconductors and internet platforms, a substantial portion of remuneration is paid not in money but in restricted stock units (RSUs), whose value depends directly on market cap.

The ratio of MCap to total wage bill in semiconductors – roughly 58 dollars of market value per dollar of salary cost – means that equity is being used as an ultra‑efficient currency to buy and retain scarce talent. A major correction in the multiple does not simply hurt investors; it instantly devalues salaries, undermines recruitment and retention, and can rapidly translate into a human capital crisis.

The Apple and Nvidia episodes unfold squarely in this regime. When Apple sheds hundreds of billions of dollars in a matter of days, the RSUs of its engineers, designers and AI researchers lose a large fraction of their real purchasing power; when the stock rebounds and Apple overtakes Nvidia as the world’s most valuable company, those same RSUs regain that power. When Nvidia’s valuation races towards 5 trillion on the back of AI demand and then pulls back sharply, the wage structures of the sector’s core talent pool are stretched and compressed along with the price. https://finance.yahoo.com/video/apples-market-cap-tops-5t-194827598.html 

Volatility of this magnitude is therefore not an abstract curiosity. It is a direct modulation of the livelihoods, career choices and migration decisions of highly skilled workers – the same engineers and researchers on whom the system depends to deliver the improbable profit trajectories embedded in current valuations. An engineer earning $300,000 a year—half in stock—sees the value of their compensation package plummet by over 40%. In such a scenario, these professionals can no longer cover their fixed expenses (which can range from $120,000 to $150,000 annually in the Bay Area).

Yet, in most commentary, these moves are still framed as “market reactions” to quarterly guidance or AI roadmaps, as if the underlying arithmetic of wages‑via‑equity and the fragility of talent‑dependent production were mere footnotes. This is a form of denial: the abyss is visible in the numbers, but the human dimension is systematically excluded from the narrative.

III. The Denial of the Abyss: Irrationality with Eyes Wide Open

The word “irrationality” has now begun to appear explicitly even in the discourse of major AI platforms. Alphabet’s CEO has admitted that the current AI investment boom contains “elements of irrationality”, echoing Alan Greenspan’s famous warning of “irrational exuberance” before the dotcom crash. Meanwhile, market participants continue to rotate trillions of dollars of value between a handful of tech giants based on small updates to AI strategies, product keynotes, or rumours of acquisitions, as if nothing in the physical or geopolitical environment had changed. https://www.bbc.com/news/articles/cwy7vrd8k4eo 

In The Loop and the Black Hole, I argued that what the market is betting on, in semiconductors and AI, is not a normal business cycle but a kind of physics experiment: the construction of a different reality, one in which AI‑driven automation, autonomous vehicles, and fully robotized supply chains replace large fractions of today’s human labour in services, manufacturing and logistics. The valuation arithmetic demands that this new reality not only be built, but built fast, with profit growth in the high double digits for two decades.

The abyss is the recognition that this transformation sits on multiple hard constraints:

• Material constraints: finite supply of cobalt, gallium, germanium, rare earths, and ultra‑pure water; concentrated production chains in the Democratic Republic of Congo, China, Taiwan, Malaysia and other nodes with significant political and climatic risk.
• Energetic constraints: data‑centre electricity demand rising from ~200 TWh to 300–400 TWh within a few years, competing directly with domestic electrification and climate‑transition goals in advanced and emerging economies.
• Fiscal and institutional constraints: states whose capacity to manage destruction‑creation cycles are eroded by chronic under‑taxation of Big Tech and by debt burdens that limit room for countercyclical policy.
• Geopolitical constraints: weaponized interdependence in semiconductor exports, rare‑earth supply, and monetary policy, where any tightening by the Federal Reserve acts as a capital vacuum for emerging markets.

The market is not ignorant of these constraints; it simply chooses not to integrate them into the pricing of AI‑linked assets in any systematic way. When Nvidia reaches 4–5 trillion of market cap and Apple oscillates around a similar level, the implicit assumption is that all of these constraints will either be magically resolved or remain sufficiently offstage to leave the profit trajectory intact.

This is not simple irrationality; it is a structured denial – a refusal to accept that the same physics, geopolitics and demography that govern the rest of the world apply also to the loop.

IV. The Loop Tightens: Apple Inside the Black Hole

In the original article, the loop was defined by a closed circuit: Big Tech platforms (OpenAI, Microsoft, Google, Meta) purchase GPUs from Nvidia at very high margins, financed by current revenues and by capital inflows premised on future AI profits; Nvidia’s market cap rises, validating the narrative; the platforms point to Nvidia’s success as proof of the wisdom of their AI investment; capital flows increase; more GPUs are bought; the loop reinforces itself.

There is little economic “spillover” to bakeries, textile factories or hospitals; most of the money circulates among the same actors, with limited diffusion into the broader economy. In recent developments, Apple is being pulled decisively into this loop: its valuation now fluctuates on the perceived quality of its AI roadmap, the integration of on‑device models, and the possibility of acquisitions in AI search or AI chip design.

When Apple is punished with a multi‑hundred‑billion drawdown for a “disappointing” AI presentation and then rewarded with a near‑5‑trillion valuation after sentiment shifts, what is being evaluated is not the iPhone, the MacBook or the App Store as such. It is the degree to which Apple is deemed to participate in the same closed gravitational field as Nvidia: a region where assets validate one another’s narratives and where every AI keynote is treated as a small perturbation in a massive black‑hole geometry.

In astrophysical terms, Apple has crossed the loop’s event horizon. Its evaluations only make sense inside the loop’s own logic. From outside – from the perspective of Kolwezi, Penang, Manila or the lithium triangle – these valuations look like anomalies: a star system emitting blinding light while drawing in matter, labour and resources from its surroundings without materially changing their development trajectories.

The denial of the abyss is therefore double: at the centre, by investors and executives who treat the loop as a semi‑autonomous universe; and at the periphery, by policymakers who hope that some of the emitted light will eventually translate into public revenue, social protection or sustainable infrastructure, despite the current arithmetic suggesting otherwise.

V. Negation as Strategy: Too Big to Fail, Too Big to Admit Limits

The original article concluded with three evaporation scenarios: stagnation of scaling laws in AI, an external technological leap that renders current GPUs obsolete, and a monetary‑energy squeeze that forces a brutal repricing of long‑duration tech assets. None of these scenarios has materialized fully – yet – but the recent behaviour of Apple and Nvidia can be read as an early rehearsal.

The system’s response so far has been to double down on denial. Central banks remain implicitly committed to a “Fed put” that keeps discount rates artificially low relative to structural risk, because a full repricing of tech would transmit directly into pension funds, insurers and systemic stability. AI platforms continue to announce ever‑larger training runs and capital expenditures, sometimes reaching verbal commitments in the tens or hundreds of billions, as if energy, materials and climate timelines were infinitely elastic.

At the same time, the political system reacts with populist reflexes that attack trade, migration or multilateralism – the very mechanisms that would be needed to manage a coordinated transition – rather than confronting the core question:

For today’s price to make sense, what kind of economy must be built tomorrow, and within which physical and political constraints?

The answer, as we have seen, is extremely demanding. It implies Schumpeterian destruction on a scale that compresses decades of sectoral transformation into a handful of years, in societies and states that no longer possess the fiscal, institutional or energy slack that previous waves of industrialization enjoyed.

To acknowledge this openly would be to admit that the current loop is, in significant measure, a bet against physics, against geopolitics, and against the demographic and institutional limits of contemporary capitalism. Instead, the system chooses negation: the abyss is visible, but the dominant narrative insists it is a mirage that can be outgrown through more scale, more models, more data centres and more trillions in market cap.

VI. A Different Kind of Rationality

Calling this situation “irrational” is both accurate and insufficient. The loop is internally coherent: within its own micro‑universe, each company’s valuation appears justified by the valuations of its peers; each AI investment seems vindicated by the latest rally in GPUs or cloud revenues; each keynote can be read as a proof point in a self‑referential story of exponential progress.

What is missing is a rationality that is external to the loop – a perspective that treats finance, AI, semiconductors, energy, climate and geopolitics as components of a single system, subject to the same finite constraints and the same non‑linearities. The Loop and the Black Hole argued that such a perspective is not optional; it is a methodological imperative.

The recent Apple and Nvidia episodes confirm the diagnosis. They are not random shocks; they are signatures of a regime in which entire trillions move on the basis of narratives that, when translated back into physical and social terms, remain profoundly under‑specified.

To refuse to see this gap is to practice a form of financial negationism: a deliberate blindness to the abyss at the centre of the loop – not the abyss of collapse, but the abyss of inconsistency between the prices we accept today and the world we are plausibly able to build tomorrow.

The task ahead is therefore not to predict the exact date of a crash or the precise path of future valuations. It is to restore an external rationality: one that insists that Kolwezi, Penang, Manila and the lithium triangle belong on the same axes as Cupertino, Santa Clara and Wall Street; that climate, energy and fiscal capacity are not footnotes to AI, but boundary conditions; and that any valuation which ignores these boundary conditions is, in the strict sense of the term, irrational – regardless of how many models it trains or how many trillions it briefly reaches.