The AI Reset: Where We Actually Are—and What Comes Next
Over the last 6–12 months, the market has aggressively repriced a large portion of the technology sector—particularly companies tied, directly or indirectly, to artificial intelligence. Many of these stocks are down anywhere from 30% to 70% from their highs.
That move wasn't random. It was Phase 1.
Phase 1: Narrative Collapse
The initial phase of this cycle was driven almost entirely by narrative. The market moved from "AI will enhance everything" to "AI will destroy existing business models."
That shift led to indiscriminate selling—good businesses and weak ones treated the same. We've seen this before: Zoom Video Communications post-COVID, Shopify after e-commerce normalization, Meta Platforms during the 2022 ad slowdown. In each case, the narrative overshot reality. That's what markets do.
Where We Are Now: Transitioning Out of the Fear Phase
We are likely at—or very close to—the bottom of the narrative-driven decline. Not because fundamentals have fully recovered, but because the uncertainty discount has already been priced in.
The market has already asked: "What if AI destroys this business?" Now it needs to answer: "Is that actually happening?"
That's Phase 2.
Phase 2 (Next ~6 Months): Fast Snapbacks
Some companies are not being disrupted by AI—they're leveraging it immediately. These are the names where AI improves margins, enhances product value, and strengthens customer retention.
A good example is Toast. Toast operates in restaurant software—payments, POS systems, operations. AI doesn't replace that. It enhances it: better demand forecasting, smarter pricing tools, operational automation for restaurants, improved analytics for owners.
If Toast shows stabilizing growth (currently mid-teens revenue growth), expanding margins, and AI-driven efficiency gains, the stock doesn't need perfection. It just needs the narrative to be less wrong than currently priced in. That's how you get sharp moves.
These types of companies can re-rate quickly—often within 1–2 earnings cycles. We've seen this exact pattern before with Meta in 2023's "year of efficiency" (+150%) and NVIDIA's exponential revaluation as AI monetization clarified.
Phase 3 (12–18 Months): Slow Normalization
The second group is more complex. These are companies where the market has legitimate questions: Will AI commoditize their product? Will pricing power decline? Does their business model need to change?
Examples across the market include design software (Adobe), developer tools, and certain SaaS platforms with seat-based pricing.
The key issue here is structural. Many of these businesses were built on per-seat pricing and human-driven workflows. AI introduces token-based usage models, automation that reduces seats, and new competitors with lower cost structures.
That transition doesn't happen overnight. It requires product redesign, pricing model evolution, proof that margins can hold—and most importantly, time for the market to trust the new model. That's why these names typically take 12–18 months to stabilize and re-rate. Not because they're broken, but because they need to prove durability in a new paradigm.
What the Market Is Missing
Right now, the market is still broadly treating "AI exposure" as a risk. That's backward. The real divide is this:
- AI as a tailwind — Enhances the business, improves margins, strengthens moat.
- AI as a disruption risk — Compresses pricing, changes user behavior, forces business model evolution.
The opportunity is in identifying which is which—before consensus shifts.
Positioning From Here
If you step back, this is not a collapse. It's a reset. And resets create opportunity—but only if you're precise.
Here's the practical framework:
- Near-term (0–6 months): Focus on companies where AI is already showing up in results. Expect faster re-ratings.
- Medium-term (6–18 months): Focus on quality businesses undergoing transition. Expect volatility, but improving clarity.
- Avoid: Companies where AI genuinely weakens the core model with no clear adaptation path.
Bottom Line
The market has already priced in the fear. What it hasn't priced in yet is who adapts successfully. That's where the opportunity is.
Some names will move quickly—because the narrative is already wrong. Others will take time—because the business itself needs to evolve. But broadly speaking, the idea that AI "destroys everything" has already run its course.
Now we move into the phase where reality—not speculation—drives returns.