Artificial intelligence could upend the pecking order in technology, BCA Research’s Peter Berezin warned, arguing that new winners may emerge as incumbents face sudden pressure on profits and control. Speaking on Fox Business’ “Making Money,” the chief global strategist outlined how AI may change who benefits most from digital growth and when that shift could show up in markets.
Berezin’s message targeted investors glued to a narrow set of mega-cap leaders. He said the same forces that have lifted those stocks could spread gains, compress margins, and redirect value to different parts of the economy. The comments arrive as companies race to deploy AI across cloud services, search, chips, and enterprise software, with spending and expectations at decade highs.
Why AI Could Reshape Market Power
“Artificial intelligence could disrupt tech giants and fundamentally reshape the balance of market power,” Berezin said.
His case rests on where profits accrue in an AI stack that ranges from chips and data centers to models and applications. In recent years, the largest platforms captured the most value thanks to network effects, distribution, and control over user data. AI could weaken those moats in three ways.
- Hardware intensity: Training and inference require vast compute, which may shift bargaining power to chipmakers and advanced manufacturers.
- Open models: If capable open systems spread, differentiation based on proprietary models may narrow, pressuring pricing and margins.
- Commoditization risk: AI features can be copied quickly, reducing lock-in and moving value to those with unique data, workflow integration, or regulation-proof niches.
Berezin suggested that profit pools could migrate from platforms to suppliers or to specialized apps that solve narrow, high-value tasks. That would mirror past tech shifts when value moved from device makers to operating systems, then to cloud platforms.
Lessons From Past Tech Cycles
Market leadership changes after major technology waves are common. The PC boom rewarded software franchises more than hardware. The smartphone era boosted integrated ecosystems. The cloud era favored hyperscalers and subscription software. Each turn saw strong incumbents stumble as costs, standards, and user behavior changed.
AI could follow a similar arc. Early gains often cluster in a few stocks. Over time, competition and regulation catch up. Profit margins normalize, and new categories appear. For investors, timing that transition is crucial.
Winners And Losers In An AI World
The near term still favors firms that control compute, capital, and distribution. Large platforms can fund model training, secure chip supply, and reach billions of users. They also own the cloud contracts where AI services run.
Yet Berezin flagged risks for those leaders. If open-source models reach near-parity for many tasks, customer switching becomes easier. If developers write once and deploy across platforms, platform fees face pushback. If regulators restrict data use, proprietary advantages shrink.
Potential beneficiaries extend outside mega-cap tech. Chip designers, contract manufacturers, power suppliers, and data-center builders could gain from the infrastructure buildout. Enterprise software vendors with deep workflow hooks may defend pricing even if models commoditize. Niche data owners could see increased licensing value.
What This Means For Investors
Berezin’s analysis points to portfolio concentration risk. A handful of stocks now drive major indexes. If AI economics broaden, that skew could unwind. Sector and style leadership might rotate.
Key signals to watch include:
- Gross margin trends at model providers and platforms.
- Chip supply and pricing, especially for leading-edge nodes.
- Adoption of open-source models in large enterprises.
- Regulatory moves on data privacy, AI safety, and competition.
He also implied that cash returns may face pressure if firms prioritize capex for compute and power. Dividend and buyback plans could adjust as companies chase model performance and energy capacity.
Counterpoints And Constraints
There are strong counterarguments. Scale still matters in AI training, reinforcement, and safety. Distribution, trust, and compliance favor established vendors. Large platforms can bundle AI into existing products, raising switching costs. Many customers prefer integrated solutions over piecemeal tools.
Energy and talent remain chokepoints. Power-hungry AI workloads may face grid limits. Hiring and keeping scarce engineers is expensive. These frictions could slow disruption and extend the advantage of incumbents with capital and partnerships.
Regulation could also cement the status quo if new rules raise entry barriers. On the other hand, antitrust scrutiny could open doors for challengers by curbing bundling and exclusive deals.
Berezin’s bottom line is that AI’s impact will not be linear. Markets may overprice early leaders and underprice second-order winners tied to supply chains and enterprise adoption.
The next phase will turn on execution, cost curves, and policy. For now, investors should test assumptions about where value accrues, diversify across the AI stack, and track leading indicators of margin pressure. The balance of market power may change faster than expected—and in less obvious places.
