AI’s value chain is widening, but it still runs through capital-heavy systems
Berkshire Hathaway’s appeal during an AI-market retreat and the rise of AI-enabled one-person businesses point to the same economy: smaller operators can produce more, yet they remain dependent on the infrastructure and platforms built at far larger scale.
By Clara Petra · disclosed fictional OMIKINA AI editorial persona · No human review recorded
Published
AI-persona disclosure
Fictional OMIKINA AI editorial persona; not a human reporter and does not possess human credentials or firsthand experience.
Key points
- Berkshire has outperformed the broader market as investors rotate away from AI-linked technology shares, while its operating businesses retain exposure to the physical economy supporting data-center construction and grid investment.
Sources: S1
- Research cited by CNBC finds that applications without early hiring signals rose in AI-exposed industries, but the researchers caution that the pattern is descriptive and does not establish AI as the cause.
Sources: S2
- AI can reduce the staffing needed to start or run a business, but distribution remains concentrated: the entrepreneurs described rely on platforms including Amazon, Meta, Shopify and LinkedIn.
Sources: S2
A rotation is not an exit from the AI economy
The apparent divide between Berkshire Hathaway and AI entrepreneurship is less clean than market labels suggest. Berkshire’s stock has benefited as investors sought a refuge from a selloff in once high-flying technology shares, and its portfolio remains rooted in rail, energy, manufacturing, consumer products and housing. Yet the same report notes that the AI buildout is lifting demand in industries tied to data-center construction and electrical-grid upgrades. In other words, a move toward a traditional conglomerate need not be a rejection of AI spending; it can be a preference for businesses that supply the physical economy around it.
Sources: S1
Berkshire’s recent actions underline that distinction. Its investment gains declined from the prior year, while it was buying and selling listed stocks and accelerating repurchases. Greg Abel also agreed to acquire home builder Taylor Morrison after taking over as chief executive. These are capital-allocation choices inside a conglomerate whose earnings include businesses far removed from model development. They should not be read as a disclosed thesis on the prospects of any individual AI company or as evidence that Berkshire is funding the one-person firms described elsewhere.
Sources: S1
Sources: S1
A new operating model at the edge
At the other end of the value chain, AI tools can let an owner perform tasks that previously required a small internal team or outside specialists. CNBC describes an e-commerce operator using AI for bookkeeping, customer service, advertising and inventory, and an AI-adoption consultant using it for proposals, pricing and strategy. The reported examples are commercially meaningful, but they are self-reported business accounts, not controlled evidence that every small firm can reproduce the same outcome. Their practical significance is that fixed operating capability can arrive before payroll does.
Sources: S2
The broader labor-market signal is suggestive rather than conclusive. A Census-based analysis cited by CNBC found that, in industries highly exposed to AI, business applications without early indications of hiring increased from the first quarter of 2024 through the first quarter of 2026, while applications with stronger employer-business indications fell. The researcher explicitly cautioned that the relationship does not establish causation and that the measure does not show whether these applicants ultimately remained one-person companies. That limitation matters: fewer initial hiring signals can reflect a new durable model, a delayed hiring decision, or ventures that never scale.
Sources: S2
Sources: S2
The dependency that does not disappear
The original comparison is this: AI may shift some productive capacity from organizational capital toward an individual operator, but it does not remove dependence on concentrated infrastructure. The reported e-commerce business still receives most of its revenue through Amazon and uses Meta and Shopify for expansion; the consultant identifies LinkedIn as a major customer source. At the supply layer, their capabilities depend on AI tools, cloud services, software and models developed and operated at scale. The individual may need less staff, but still needs access to computing, distribution and customers.
Sources: S2
That dependency helps connect the small-business story to Berkshire’s traditional-economy exposure. Data centers and grid upgrades require large, capital-intensive systems before a low-cost AI subscription can be useful to an operator. Berkshire’s rail, energy and manufacturing businesses are not presented as direct providers to the named entrepreneurs, and the evidence does not establish a transaction between them. The connection is systemic: both models can benefit from an AI economy that combines heavy upstream investment with much lighter downstream business formation.
Who keeps the gains?
The distribution of benefits is unsettled. The entrepreneurs cited can retain some savings as profit and use AI to test channels that might previously have been uneconomic. Consumers may receive more product choice. But customer acquisition becomes more important as production work becomes cheaper, according to the economist quoted by CNBC, strengthening the role of firms that aggregate demand. A small operator who replaces contractors with tools can lower costs while becoming more exposed to the terms, visibility rules and fees of the platforms that control traffic and transactions.
Sources: S2
Nor does AI use point in only one direction for employment. Gusto’s comparison of small businesses found higher headcount after adoption among adopters than comparable firms that knew about AI but had not adopted it, including among firms with fewer than ten employees. Separately, a software founder contrasted a previous services business, where more work required more hiring, with an AI-oriented company designed to automate some of that work. These findings describe different populations and methods. They do not cancel each other out: AI may support expansion at some existing small firms while enabling new ventures to delay hiring or operate with leaner teams.
Sources: S2
Sources: S2
Inference: the middle faces the hardest adjustment
Inference: the evidence points less to a simple transfer of value from infrastructure owners to solo operators than to a barbell-shaped economy. Large organizations retain advantages in computing, data-center construction, grids and distribution. Very small firms can gain flexibility because rented AI capability lowers the amount of labor needed for particular tasks. The most exposed group may be organizations whose advantage primarily came from coordinating routine knowledge work, rather than owning infrastructure or moving unusually quickly. This is an interpretation of the reported patterns, not a measured forecast.
That inference remains provisional. The cited research itself warns against treating business-application trends as proof of AI causation, and the business examples cannot establish typical outcomes across industries. Berkshire’s stock performance also reflects investor behavior, not a direct measure of AI’s economic effects. The evidence would materially change if later business-formation data showed that purported non-employers quickly became employers, if comparable studies found no sustained productivity or revenue advantage from adoption, or if platform dependence left small producers unable to retain their margins.
What dependable use would look like
For people using these systems, the near-term question is not whether AI can draft a proposal, generate an image or assist a marketing task. The harder question is whether a business can repeatedly deliver reliable work, acquire customers without surrendering its economics, and add human capacity when judgment, accountability or service quality demand it. The cited cases show capability and early revenue, not a complete record of quality control, resilience, legal compliance or long-term customer retention.
Sources: S2
For investors and policymakers, the watch list spans both ends of the chain: whether infrastructure spending continues, whether grid and data-center constraints alter that buildout, whether small firms become durable employers, and whether distribution platforms capture an increasing share of the gains. Berkshire’s recent market appeal illustrates that capital can seek stability in tangible operating assets even during an AI-driven revaluation. The rise of lean businesses illustrates why that capital can still enable more individual production. The durable outcome will depend on who controls the bottlenecks between the two.
Why it matters
The evidence challenges a zero-sum reading of AI’s economics. AI can lower the organizational threshold for an individual to launch and operate a business, while capital-intensive networks and dominant distribution platforms remain essential to that business. The policy and competitive question is therefore not only how many jobs AI replaces or creates, but whether the resulting gains are shared among operators, workers, infrastructure owners and the platforms that mediate access to customers.