Three Years After ChatGPT: What Investors Should Do Now
Three years since ChatGPT's public release, the AI boom has done more than push stock prices higher. It has rewired workflows, reshaped staffing, and sparked a massive infrastructure buildout. The result: a distinctly K-shaped economy with wider gaps between the corporate and consumer winners and everyone else.
What makes this cycle stand out is the context it came from. The market was bruised, sentiment was shot, and tech was the epicenter of the damage. Then a six-sentence product announcement lit the fuse.
The 2022 Setup That Made the Upside Possible
On Oct. 12, 2022, the S&P 500 bottomed, down about 25% from its January peak. By Nov. 30, when ChatGPT launched, the index had bounced roughly 13% off the lows but wouldn't reclaim a new high until January 2024.
Inflation was hot. The Fed was hiking aggressively. Market darlings took the hit: Nvidia, Meta, and Palantir fell as much as ~70% in 2022. Apple slid ~30%, Alphabet ~40%, and Amazon was cut in half. Skepticism about any new bull run was rational.
From Six Sentences to a Full-Blown AI Cycle
OpenAI's announcement was brief: a conversational model that could answer follow-ups, admit mistakes, challenge incorrect premises, and refuse inappropriate requests. That was it. No grand vision. No glossy keynote.
Yet the impact was outsized. Eighteen months before launch, OpenAI's private valuation sat near $14 billion; today it's reported around $500 billion. Sam Altman became the face of a product that converted mass curiosity into enterprise spend.
Why It Mattered for Markets
- New profit pools: AI shifted the earnings narrative from ad-driven growth to compute, models, and applied software ROI. Investors had something tangible to underwrite again.
- Capex supercycle: Hyperscalers and enterprises opened the checkbook for GPUs, data centers, and power. The spend spilled into semis, equipment, real estate, and utilities.
- Productivity re-rated: Teams kept headcount flatter while output rose. Margins held up better than expected, especially in companies that integrated AI into core operations.
- Dispersion increased: The K-shape widened. Winners compounded network effects and data moats; laggards faced rising cost to compete.
- Index concentration: Mega-cap AI leaders carried benchmarks, while breadth lagged. Stock picking started to matter again.
What To Watch Next (for PMs, CFOs, and Analysts)
- Data center bottlenecks: Power availability, lead times, and interconnect constraints will dictate build velocity and pricing.
- Unit economics of AI products: Track gross margins after inference costs, latency trade-offs, and model-switching behavior.
- Earnings attribution: Separate AI-fueled revenue from bundling/price hikes. Look for cohort retention and uplift, not just pilots.
- Opex mix and headcount: Watch for reallocation toward AI infra, MLOps, and automation offsetting traditional roles.
- Regulation and IP: Liability, training data rights, and safety standards could shift cost curves and timelines.
- Second-order demand: Semicap, networking, power, cooling, and real estate-who benefits as the stack scales?
Portfolio and Strategy Checklist
- Barbell positioning: Own "picks and shovels" (compute, memory, networking, power) on one side; targeted application winners with clear payback periods on the other.
- Cash flow discipline: Reward companies that tie AI spend to customer outcomes and durable margins. Discount hype without revenue proof.
- Energy exposure: AI demand pulls on electricity, renewables, storage, and grid modernization. Model power availability into growth assumptions.
- Diversify beyond the obvious: Look for overlooked enablers-software tooling, data platforms, security-where valuation hasn't fully priced scale.
- Risk controls: Hedge concentration risk and rate sensitivity. Scenario test slower node transitions or supply delays.
The Quiet Start Still Matters
The AI cycle began with an understated launch and grew on results, not rhetoric. That pattern will repeat: lower-cost compute, better models, and clearer business cases feeding steady capex and cash flows.
Keep your process simple: follow the earnings, test the unit economics, and ride the compounding where the infrastructure and adoption curves intersect.
Practical Resource
- Curated AI tools for finance teams - shortlist solutions worth piloting inside FP&A, research, risk, and IR.
Your membership also unlocks:
