Crossing the AI Chasm: Why the Real Profit Wave is Slated for 2028

In a recent compelling interview with CNBC, Dan Chung, CEO and Chief Investment Officer of Alger, shared a perspective that cuts through the current market noise: we are not in an AI bubble; we are in the "road-building" phase of a generation-defining infrastructure build-out.

While many investors are sweating over the massive capital expenditures (CAPEX) from Big Tech, Chung argues that for the patient investor, this isn't a red flag—it’s a generational buying opportunity.

The "Crossing the Chasm" Framework

Chung frames the current AI landscape using Geoffrey Moore’s classic "Crossing the Chasm" model. We have moved past the early adopters and are now in the difficult transition toward the early majority.

The "chasm" exists because while the technology has proven its concept, the physical infrastructure—data centers, power grids, and high-speed fiber—isn't yet robust enough to support mass commercialization. This is why we see a delay between massive spending and massive earnings.

The 7000 Billion-Dollar Bet

The numbers are staggering, but Chung provides the necessary context. The "Big Four" (Amazon, Google, Meta, and Microsoft) are projected to spend roughly $7000 billion on AI infrastructure.

• Amazon is funneling 80% of its $200 billion budget into AWS data centers and specialized AI campuses.
• Microsoft is deeply integrated with OpenAI, scaling computing power at an unprecedented rate.

Chung predicts a pivotal "Inflection Point" in 2028. Currently, CAPEX growth is screaming at 60–70% annually. By 2028, as the "roads" are finished, this spending will rationalize to 20–30%. When that happens, the Free Cash Flow (FCF) will revert to shareholders like a tidal wave.

"The market is a forward-looking machine," Chung noted. "Wall Street will begin pricing in this 2028 cash flow feast as early as 2027."

Beyond the "Mag 7": Where the Real Alpha Lives

While Chung remains constructive on the Magnificent Seven, he highlighted two specific sectors where the AI revolution is creating immediate supply-demand imbalances:

1. The Semiconductor & Memory Squeeze

AI training requires HBM (High Bandwidth Memory). The demand is so high that the HBM market is expected to grow from $35 billion in 2025 to $100 billion by 2028.

• The Reality: Capacity for 2026 is already sold out.
• The Opportunity: Because manufacturing plants take years to build, traditional server memory prices have surged by 170%. Chung pointed to Western Digital as a primary example of a company whose enterprise capacity has been effectively "cannibalized" by AI demand.

2. Healthcare: The Non-Tech Beneficiary

Perhaps the most exciting part of the interview was Chung’s pivot to Healthcare. AI is transforming drug discovery from a "hit-or-miss" 15-year cycle into a precision 3-to-5-year process.

• Efficiency: AI can cut drug discovery costs by 70%.
• Valuation: The healthcare sector (XLV) has been underperforming due to policy uncertainties, creating a "perfect storm" of low valuation, an aging global population (1 billion people over 65 by 2026), and AI-driven innovation.

The Verdict: A Pullback is a Gift

Dan Chung’s message to the market is clear: Don’t confuse heavy investment with a lack of returns. The productivity gains are already showing—2026 data indicates a 11.5% average productivity boost for companies using AI for over a year.

For those watching the daily charts, the message is simple: The infrastructure is being built, the "chasm" is being crossed, and the real harvest of the AI era is just over the horizon.