Is Generative AI Becoming Foundational Technology?

Lanre Adeoye

Image credit: Canva

As I learn more about AI and its implications, the fundamental question that keeps surfacing for me is this: is generative artificial intelligence, or Gen AI, emerging as a foundational, general-purpose technology in the same way electricity or the internet once did?

At first glance, that comparison can feel overstated. AI is not new. It is built on decades of computer science and depends heavily on the internet’s infrastructure. But foundational technologies are not defined by novelty. They are defined by the patterns they create over time. When I examine the trajectory of Gen AI, I see dynamics that resemble the early arcs of electricity and the internet more than those of incremental software tools.

Across history, transformative technologies share common markers. They reshape how value is created, how people learn, how information flows, and how societies organize trust. They demand new infrastructure, give rise to new industries, and force cultural renegotiation. Increasingly, Gen AI appears to be exhibiting these same characteristics.

Three overlapping trend lines stand out: value creation, infrastructure, and culture.

From Information Distribution to Cognitive Distribution

The internet reshaped value creation by lowering the cost of accessing and distributing information. Knowledge became abundant, collaboration global, and coordination cheaper.

Gen AI follows a similar trajectory, but with an important difference. Rather than distributing information, it distributes cognitive capability.

When software can draft, summarize, reason across domains, write code, analyze data, and adapt outputs dynamically, the unit of leverage in knowledge work shifts. Work becomes less about retrieving information and more about directing, validating, and compounding cognitive output. Expertise is no longer applied only where it exists; it can be instantiated, amplified, and scaled.

This is also where skepticism is strongest. Critics argue that these systems do not truly reason and may eventually hit a ceiling. If Gen AI is ultimately sophisticated pattern matching rather than cognition, its impact could be limited.

But foundational impact does not require human-like intelligence. Electricity did not “understand” mechanics, yet it reorganized industry. The internet did not comprehend meaning, yet it transformed communication and commerce. What matters is not how AI thinks, but whether societies reorganize around what it can reliably do. Even derivative cognition, when externalized and scaled, alters economic structure.

For founders and investors, this suggests a shift deeper than automation. Productivity gains are not confined to isolated workflows; they affect team design, product cycles, and the speed at which organizations learn and adapt.

Infrastructure Signals: Bigger Than Applications

Foundational technologies are never just user-facing tools. The internet required cables, data centers, protocols, cloud infrastructure, and new professional classes.

Gen AI is already pulling on similar layers. Demand for compute is reshaping semiconductor roadmaps. Energy availability is becoming a strategic constraint. Data pipelines, deployment tooling, observability, security, and governance are professionalizing rapidly. Regulatory frameworks are forming alongside them.

These signals can be read as both momentum and risk. Electricity and the internet became foundational because they grew cheaper and more ubiquitous. If Gen AI remains expensive or environmentally untenable, it could remain concentrated rather than general-purpose.

That risk is real. But historically, infrastructure-heavy technologies look least efficient early on. Early electricity grids and early internet access were expensive, uneven, and fragile. Over investment and competition eventually drove down costs. The current AI infrastructure race resembles this early, wasteful phase more than a stable ceiling.

The Bubble Makes the Point

It is common to describe the current AI moment as a bubble. Capital is moving faster than fundamentals. Many products will fail.

This is not a refutation. The dot-com bubble did not negate the internet’s importance; it accelerated it. Infrastructure was overbuilt, expectations collapsed, and many firms disappeared. Yet the underlying capabilities became cheaper, more reliable, and deeply embedded in the economy.

The same pattern appears to be unfolding with Gen AI. Volatility is concentrated at the application layer, where differentiation is thin and switching costs are low. Many AI-native startups will not survive. But beneath that churn, foundational layers; compute, data infrastructure, deployment, safety, and governance, continue to compound.

From this perspective, bubbles are not evidence against foundational technologies. They are often the mechanism by which societies overinvest early and accelerate long-term adoption.

Cultural Renegotiation and Trust

Every foundational technology disrupts trust systems. Exams worked because knowledge was scarce. Credentials worked because skill acquisition was slow and legible.

When assistance becomes ambient, those systems break down.

Gen AI is forcing institutions to renegotiate how competence is validated. We are already seeing shifts toward supervised exams, oral defenses, project-based assessment, continuous evaluation, and alternative credentials. This is not incidental. It is a cultural response to a world where knowing and doing are no longer easily separable.

Here, another risk emerges: data degradation. Gen AI depends on high-quality human data. If models increasingly train on AI-generated content, feedback loops could degrade reliability.

This challenge is unresolved. But it mirrors earlier transitions. Industrialization polluted before regulation emerged. The internet degraded information quality before new norms and institutions formed. Foundational technologies generate second-order problems precisely because society reorganizes around them.

What Would Have to Be True for Gen AI Not to Become Foundational?

For Gen AI not to become foundational, several conditions would likely need to hold simultaneously. Scaling would need to hit hard limits. Regulation or social rejection would need to prevent deep institutional embedding. Trust failures would need to outweigh productivity gains at scale.

None of these outcomes are impossible. But together, they require multiple trend lines to reverse direction at once. At present, those lines continue to point toward deeper integration.

Why This Matters Now

Foundational technologies rarely reveal their full impact early. Electricity existed long before it reorganized factories. The internet existed long before it reshaped commerce and media.

Gen AI may still be early. But the economic, infrastructural, and cultural signals are already visible. The question is not whether AI will ship better tools, but whether we are witnessing the emergence of a new general-purpose layer; one that redistributes cognitive capability the way electricity redistributed mechanical power.

Conclusion

If Gen AI is becoming foundational, then the most important opportunities are not at the application edge alone.

They are in businesses that assume cognitive capability will become cheap, abundant, and embedded, and build systems accordingly. That means investing in infrastructure, workflows, governance, and institutions that are resilient to model churn and application turnover. It means designing organizations that treat AI not as a feature, but as a baseline capability. And it means asking not “what can AI do today?” but “what becomes possible once cognition itself is no longer scarce?”

If that assumption is wrong, many bets will fail. But if it is right, the cost of ignoring it will be far higher.

About the Author

Lanre Adeoye is a talent and business operations leader with experience at the intersection of people, technology, and strategy. An MBA graduate of London Business School, she has helped startups and multinationals scale across regions through innovative approaches to recruitment, organizational design, and workforce transformation. Her work now explores how AI and emerging technologies are reshaping work, leadership, and venture growth across industries.

Say hello on LinkedIn or at lanre.a@workarena.co