Will AI Really Change the World? Capital Group Maps Four Possible Futures
Economist Jared Franz has drawn a scenario map of what may lie ahead: from a triumphant super‑cycle to a bubble burst, with a more uneven path in between. The crucial question is no longer whether AI will succeed, but how fast and under what conditions it will do so.
It is no longer a matter of “if.” Artificial intelligence is already embedded in companies, woven into workflows, and reflected in the balance sheets of chip manufacturers and data‑center operators. The question that keeps investors—and policymakers, managers, and anyone concerned with the direction of the global economy—wide‑awake is another one: which of the possible futures are we heading toward?
Franz attempts to answer this by presenting an analysis built around four scenarios that chart the AI landscape along two fundamental axes:
Degree of technology adoption in the real economy
Stance of government and financial policy – accommodative vs. restrictive
Four quadrants, four destinies. All are plausible and, to varying extents, already under way.
Scenario 1 – The Super‑Cycle: Everything Works in Harmony
In this best‑case scenario all the pieces click together. AI becomes a cross‑sectoral infrastructure: firms reorganize processes, automate repetitive tasks, and use AI tools not just to do more, but to do things differently. Computing costs fall, lowering the entry barrier. Governments back the transition, aware of its strategic importance in global competition. Productivity gains translate into higher profits, which fuel more investment, which in turn drives fresh productivity—a virtuous, self‑reinforcing loop.
The outcome is a prolonged period of robust growth and expanding margins—an era, not merely a cycle. Franz calls it “capital deepening”: a structural rewrite of how the economy functions rather than a temporary rebound.
Early signals that we are moving into this quadrant are already visible: hyperscaler capex is at historic highs, data‑center construction is booming, and the first AI‑linked productivity metrics are beginning to appear in macro‑economic data.
Scenario 2 – The Balanced Path: Forward, but in Stutters
The second scenario is more modest—and perhaps more realistic. AI advances, but unevenly. Some firms sprint, others jog, and a few still hesitate to get up from their chairs. The reasons are concrete: lingering high financing costs in certain sectors, legacy IT systems that are hard to integrate, evolving regulations that generate uncertainty, and mixed political signals across jurisdictions.
The trajectory resembles a staircase rather than an escalator: progress comes in steps, with pauses and asymmetries. Sectors such as finance, pharmaceuticals, and technology move swiftly, while others wait for clearer conditions. The result is an economy with differentiated speeds, delivering real but selective returns, and intensifying pressure on investors to separate winners from laggards.
Scenario 3 – The Bubble Burst: Investment Outpacing Reality
The third scenario is the one no one wants, but history knows well. Capital precedes returns. Financial conditions tighten—higher rates, tougher credit, and governments forced to re‑prioritize spending. Regulatory scrutiny intensifies, especially around data security, market concentration, and employment impact. Some data‑center projects stall; certain semiconductor supply chains prove over‑built.
Franz stresses that the critical point is not that AI vanishes, but that investment exceeds the underlying economic base. Companies reassess deployment timelines, investors shift toward stability, and the previously generous valuation narrative loses touch with fundamentals. We have already seen similar patterns with the late‑1990s Internet boom and the renewable‑energy surge a decade ago. It would not be the first, nor likely the last, technology bubble.
Scenario 4 – A Return to the Pre‑ChatGPT World: AI as a Tool, Not a Revolution
The fourth scenario is the quietest—and arguably the most insidious—because it does not announce itself with a crash but with a gradual dimming of expectations. AI is adopted, but only marginally: a few upgraded dashboards, partially automated workflows, modest process tweaks. A radical transformation never materializes. Companies experiment without full commitment, hampered by fragmented data, rigid organizations, and difficulty scaling innovation.
In this world, liquidity sustains markets, feeding narratives more than production. Valuations drift away from genuine efficiency gains. Growth continues to rely on traditional drivers, with AI remaining a decorative element rather than a macro‑economic lever. Not a bursting bubble, but a fire that slowly goes out.
Where We Stand – What to Watch
Capital Group’s analysis does more than outline four possibilities; it uses them as a compass. Today that compass points toward a constructive quadrant: productivity gains are becoming visible in the data, infrastructure investment remains high, and the policy environment in major economies is largely innovation‑friendly. Yet the picture is not uniform, and the distance between the super‑cycle and the balanced path—or between the latter and a bubble burst—will hinge on variables still taking shape.
For investors, the practical lesson is not to bet everything on a single scenario, but to learn how to read the transition signals:
Speed of real‑core AI integration across firms
Evidence of durable, not just episodic, productivity improvements
Trajectory of hyperscaler capex
Regulatory responses in the coming quarters
AI is advancing rapidly; the economy, as always, will adjust more gradually. The gap between these two paces is today’s primary risk—and, simultaneously, the main opportunity.