The Long Run Belongs to Technology. Does It Belong to Tech Stocks?
The strongest case for technology is not that a crash cannot happen. It is that innovation, adoption and productivity can keep creating valuable businesses long after today’s market debate has moved on. The harder question is what investors pay for that future.
Computing keeps getting cheaper, software reaches more people, and new tools can make businesses more productive. Artificial intelligence has added a new chapter—and renewed the debate over whether the boom will last or crash.
A long-term investor need not know the date of the next market fall. The key distinction is between a promising industry and a promising investment price. A company can transform its market and still disappoint shareholders who paid too much for that future in advance.
The case for a long technology cycle
Technology has a long record of moving from specialist tool to ordinary infrastructure. Computing power, connectivity, cloud services and mobile devices became embedded in work and daily life. The next wave may be AI systems that interpret information, generate content, and make constrained decisions inside existing software. If that shift spreads, the beneficiaries could include chip designers, manufacturers, cloud platforms, software companies, data-centre suppliers and businesses that use the tools well.
The economic mechanism is straightforward. If a company can serve more customers, complete a task faster, or reduce the cost of a process, it may create more value from each employee or unit of capital. Providers that own scarce infrastructure or products customers rely on may capture part of that value. The gains will not be evenly shared: customers may keep much of the benefit through lower prices, while competition pushes down suppliers’ margins. Productivity growth is therefore a reason to study technology companies, not proof that every technology share will rise.
Current company results show that this is more than a distant possibility. Nvidia reported $96.2 billion in revenue in its second quarter of fiscal 2027, including $89 billion from data centres. Broadcom said its third-quarter fiscal 2026 revenue rose 86% year over year, with AI semiconductor revenue up 221%. TSMC reported second-quarter revenue of $40.2 billion and said it expected 2026 revenue growth slightly above 40% in US-dollar terms. These are powerful signs of demand across chips and manufacturing, but they are snapshots, not promises through 2035. (Nvidia; Broadcom; TSMC)
Microsoft sells established software and cloud services while investing heavily in AI infrastructure. Its cloud business may benefit if companies rent more computing capacity and buy AI features within products they already use. That installed customer base may make adoption easier. But the investment has to earn its keep: Microsoft has described AI infrastructure costs and a shift in revenue mix as pressures on cloud margins. (Microsoft FY2026 Q4 results; earnings call)
Apple offers another exposure: its device ecosystem and services could benefit if new features persuade customers to stay or upgrade. Its prospects do not depend on the same AI infrastructure cycle as Nvidia’s.
The investment case has two engines
Long-term share returns depend on more than revenue growth. A useful mental model is: earnings growth + cash returned to shareholders + change in valuation. A business can grow quickly while its share price goes sideways if investors had expected even faster growth. A company with steadier growth can produce attractive returns if its price is reasonable and it compounds cash flow for a long time.
The market can be right about AI’s importance and wrong about some AI stocks. Chip orders may reflect durable demand, but also purchases brought forward or spending before customers know their returns. More efficient models could expand adoption while reducing computing cost per task. A growing market does not reveal which supplier captures the best margins.
The leaders can change. In earlier technology waves, infrastructure owners, device makers, software vendors and new entrants took turns capturing value. Today’s firms have real advantages, but open models, custom chips, regulation or changing customer preferences can shift profits across the value chain.
Three paths to 2035
1. A durable expansion. AI and other technologies become routine tools across industries. Companies see measurable returns, renew infrastructure, and build new products. Spending broadens into software, devices, networking, power and industrial automation. Many technology businesses grow for years; stronger investments convert growth into durable free cash flow.
2. Growth with pauses and a valuation reset. Adoption continues, but companies take longer to turn pilots into profitable use. Data-centre spending overshoots near-term demand; margins tighten; investors lower the premium they pay for growth. Technology earnings can keep rising while share prices fall sharply or remain flat for an extended period. This happens when the business improves but the starting valuation was too demanding.
3. A sharper bust. A recession, financing shock, disappointing returns on AI spending or geopolitical disruption triggers cancellations and lower risk appetite. Chip and infrastructure suppliers feel the slowdown first; expensive growth shares can fall especially hard. Some projects are delayed, weaker providers fail, and the industry consolidates. A crash would not prove technological progress had ended, nor would every company recover.
No one can assign reliable probabilities to these paths with precision. The long-run case is strongest when it can survive all three: own businesses with financial staying power, avoid dependence on one narrow forecast, and keep enough diversification that a technology drawdown does not decide the fate of the whole portfolio.
A practical way to invest in the thesis
For many investors, the sensible starting point is broad equity exposure, then a deliberate and limited allocation to technology. That captures some upside without requiring a forecast about which chipmaker, cloud provider or AI application will dominate. Investors who choose individual companies can look for customer retention, pricing power, recurring revenue, cash generation and returns on investment—not just impressive growth rates or promises.
A diversified technology basket might span different roles: Microsoft for software and cloud; Nvidia and Broadcom for AI compute and connectivity; TSMC for advanced manufacturing; and Apple for consumer devices and services. These names still share interest-rate, sentiment, supply-chain and technology-spending risks. Familiar technology shares are not automatically diversified across the economy.
Private companies such as TypeSafe AI belong in a separate category. TypeSafe is developing Jev, a system intended to return structured decisions for software rather than conversational text. Its investor DCVC announced it led a $40 million seed round. That is evidence of venture backing, not proof of repeatable demand or that an outside investor can buy shares on attractive terms. Private investments add liquidity, valuation, dilution and access risks. (DCVC announcement; TypeSafe AI)
The same discipline applies to SpaceX or any other company that is private, newly listed, or available only through a secondary offer: verify the security, ownership rights, price, fees and route to liquidity before treating it as an investable share. A famous name is not a substitute for clear terms.
Investors can spread purchases over time, rebalance oversized positions, and decide in advance what loss they can tolerate. These steps cannot prevent a drawdown, but may reduce the chance that one difficult month drives a permanent decision.
The long run is an opportunity, not a timetable
Technology is likely to remain a powerful force through the medium and long term. It is reasonable to invest for that possibility. It is not reasonable to infer that today’s leaders must keep outperforming, that prices can only rise, or that a crash is either certain or impossible.
The more useful question is: if adoption continues, which businesses can turn it into cash flow, and what price gives me a fair chance of sharing in that value? Then ask the opposite: if growth disappoints for several years, can I hold this position without damaging the rest of my plan?
A measured investor can be optimistic about technology and cautious about the market at the same time. Innovation creates opportunity, while valuation and risk determine what an investor actually earns.
This article is for general information, not personal investment advice. It presents scenarios rather than forecasts. Company results and management outlooks are dated disclosures; share prices and valuation multiples change continuously. Verify current prices, valuation, portfolio fit, taxes and fees before investing.
Originally published on Substack on 23 September 2026.