Artificial intelligence is scaling faster than any technology in history. Trillions are pouring into models, chips, data centers, cloud infrastructure, and energy systems, fueling an unprecedented global economic buildout. Global corporate investment has doubled in the past year, and total spending on AI data centers can reach $7 trillion by 2030.
Yet the ultimate test is not the size of investment — it’s the reality of return.
A stark disconnect has emerged. While 88% of businesses worldwide have adopted AI, the vast majority see no significant bottom-line impact. The AI value paradox is loud: massive investment and widespread adoption, but highly elusive returns.
The AI industry itself may have sent the wrong message by framing AI primarily as a tool for quick productivity gains and cost savings. The bottleneck is not technology; it is strategy and implementation. The true economic engine lies in core business transformation. That’s what history tells.
The Internet Precedent
The closest historical parallel is the birth of the commercial internet. It did more than speed up communication or reduce transaction costs. Digital commerce expanded markets and created new categories of products and services. Non-viable trades disappeared, but new sectors and occupations became possible.
Today, a growing consensus is emerging among experts. “Using AI to boost productivity is unlikely to create a sustainable advantage,” McKinsey points out in a recent report. “The real value from AI will lie in reshaping offerings, business models, and market structures before competitors do.”
The Sector Split: Performance vs. Core Economics
Opportunities vary across the corporate landscape, but the rule remains the same: the closer AI moves to the core of the business, the greater the potential payoff.
In knowledge sectors like legal and consulting, AI is poised for the fastest outcomes — enhancing individual performance by automating content-related tasks. Yet, contrary to early predictions, AI has not eliminated the need for human expertise. More importantly, improving task-level efficiency does not automatically translate into firm-level ROI.
The real structural shift happens elsewhere. For manufacturers and industrials, AI can alter the core business economics. Such applications include optimizing production processes, refining product design, deploying predictive maintenance, and improving quality control. The challenge is that these capabilities require complex integration and take years to develop.
The same dual reality applies to high-stakes fields. In finance and insurance, AI helps extract core value by transforming underwriting, risk management, and product innovation. The opportunities are vast, but so are the regulatory and implementation barriers.
In biotech and materials science, the opportunity is greater than ever, as AI can accelerate drug discovery, molecular simulation, and materials development. Unlike current enterprise applications, these breakthroughs depend more on specialized AI systems capable of modeling the underlying physics, chemistry, and biology of the real world. Many of these systems remain in early stages of development.
From Automation to Revenue to Growth
Early success stories showcase that value follows transformation. “Growth, not just productivity, separates AI leaders,” PWC concluded in their 2026 ROI from AI report, determining that 20% of companies capture 74% of AI-driven returns. Organizations that use the technology to transform core business processes, create new products, and improve customer interactions ultimately drive superior financial results.
Consider the success of Fifth Dimension, a real estate intelligence firm that used generative AI to dramatically reduce transaction processing times, scaling its volumes from millions to billions of dollars and expanding its market reach. Shipping giant Maersk deployed predictive AI and machine learning for route optimization and predictive maintenance to rewrite its fleet economics, saving $100 million annually. Even in customer service, ride-hailing company Lyft reduced resolution times with agentic AI by 90%, freeing up resources to reinvest in core operations.
In every case, the victory did not solely come from automating back-office tasks. It came from upgrading activities directly tied to revenue generation, operational performance, or strategic growth.
The hard truth for leaders is that extracting value at scale is exponentially more difficult than deploying AI in a few isolated functions.
The Next Phase
Massive infrastructure spending does not guarantee financial returns. Companies must actively create that value. While the first phase of AI adoption focused heavily on peripheral efficiency, the next phase will determine whether the technology will become genuinely transformative. One reduces costs at the margin, the other drives growth.
While both can move the needle, only the latter can justify the trillions.