How you’ll be using AI in 2026: note-taking rings and robotaxis

The piece outlines a 2026 inflection for AI-driven hardware — from wearable AI rings to Waymo and Tesla robotaxis — while flagging significant market and policy risks. It highlights concentration in the ‘Magnificent Seven’ (which added $3.2tn in 2025), Michael Burry’s allegation that hyperscalers have extended depreciation on expensive Nvidia chips leading to double-digit overstated earnings and tens of billions of assets at risk of write-down, and the prospect of mega-deals such as a potential $1.5tn SpaceX IPO (private trades ~ $677bn). Investors should weigh rapid product innovation and private-market valuations against increasing regulatory backlash (data-centre and social-media limits), shifting capital away from climate tech, and heightened downside risk to overvalued tech equities and dependent startup ecosystems.

Analysis

Market structure: AI-hardware and edge-device winners will be concentrated — GPU and advanced-node suppliers, select cloud infra vendors and Waymo/BIDU/UBER for robotaxi rollouts — while hyperscalers that capitalize GPUs on their balance sheets (MSFT, META, AMZN) face margin and asset-risk pressure if depreciation reversals occur. Faster chip refresh cycles create acute demand for cutting-edge silicon but shorten effective asset life, increasing replacement CAPEX and compressing secondary-market pricing within 12–36 months. Cross-asset: expect equity implied vol to rise (VIX +50–100% on a Mag7 drawdown), safe-haven bid into 10Y Treasuries and USD strength; industrial commodities for fabs (copper, polysilicon?) see lumpy spikes tied to capex timing.

Risk assessment: Tail risks include regulatory shocks (under‑16 social bans, data‑centre moratoria) and a cluster of hyperscaler write‑downs of 10–30% of stated GPU asset values within 3–12 months after audits — a scenario that could shave 15–25% off the S&P 500 via concentration effects. Hidden dependencies: local power/water permitting and TSMC/ASML delivery windows; second‑order labour displacements risk political backlash that slows data‑centre approvals. Key catalysts: quarterly filings, SpaceX IPO timeline, major data‑centre local votes and Tesla/Waymo commercialization milestones.

Trade implications: Tactical plays should be asymmetric: modest, hedged longs in NVDA/semicap supply chain (1–3% net), event short exposure to hyperscalers via 3–6 month puts sized to risk budget (0.5–1.5%), and a 2–4% allocation to long-duration Treasuries or put protection as macro hedge. Prefer pairs (long GOOGL vs short MSFT) to isolate execution/asset-accounting risk; selectively long UBER (robotaxi optionality) for 6–18 months. Use option structures to cap downside while retaining upside.

Contrarian angles: Consensus underestimates consolidation: hardware startups will fail at scale, concentrating pricing power in a few incumbents (TSMC/ASML/NVDA) — so infrastructure suppliers may be underpriced versus consumer hardware hype. The market may over-penalize NVDA on Burry-style narratives; downside risk is real but likely capped by structural GPU scarcity and long backlog through 2026. Historical parallel: 2000–03 telecom bust — infrastructure survivors captured outsized returns; expect similar outcomes here if one focuses on cash‑flowing infra names rather than consumer hardware narratives.

Key Decisions for Investors

• Establish a 1.5% notional long position in NVDA via 18–30 month LEAP call spread (buy 24‑month call, sell higher strike 24‑month call) targeting asymmetric upside to capture GPU demand; hedge with a 0.3% purchase of 3‑month 10% OTM puts to cap short‑term drawdowns. Exit/reevaluate if NVDA sells off >30% or backlog visibility deteriorates in two consecutive quarters.
• Initiate a tactical 0.75–1.25% notional short‑bias on hyperscalers (MSFT or META) using 3–6 month 8–12% OTM puts ahead of next quarterly filings to profit from potential depreciation disclosures; scale up only if an auditor/capex disclosure confirms extended useful‑life changes or if share price gaps >15% on the news.
• Run a relative‑value pair: long GOOGL (2.0% net) vs short MSFT (1.5% net) for 6–12 months to capture better monetization/AI stack control in search/ads; rebalance if spread moves >10% or after major regulatory announcements.
• Allocate 3% of portfolio to defensive macro hedges: buy 10‑year Treasury duration (or TLT equivalent) and/or 3–6 month put protection on the equity portfolio; trigger full hedge deployment if Mag7 index falls >15% within a 14‑day window.