The scientist that helped create AI says it’s only ‘a matter of time’ before every single job is wiped out—even safer trade jobs like plumbing

Yoshua Bengio, a Turing Award-winning AI pioneer, warns that AI is already displacing 'cognitive' desk jobs—hitting junior/Gen Z hires hardest—and expects broad job impacts within five years as firms automate roles; major employers including Intel, IBM and Google have frozen thousands of prospective new positions. Bengio, who has founded the AI-safety non-profit LawZero, also cautions that unchecked AI deployment could threaten democratic stability within two decades and urges corporate leaders to coordinate on risk mitigation.

Analysis

Market structure: Rapid AI adoption favors capital-intensive providers of training/inference compute, data platforms and workflow automation (NVIDIA-style GPUs, hyperscaler cloud) while compressing entry-level labor demand and legacy CPU/server vendors. Expect pricing power for accelerators and cloud (20-40% incremental margin capture at hyperscalers over 12-36 months if AI workloads scale) and margin pressure on traditional services and PC/server OEMs as hiring freezes and automation cut GFY operating costs.

Risk assessment: Tail risks include swift regulatory limits on model size/data access, export controls on accelerators, or political backlash that curtails SaaS monetization—each could erase >30% forward EBITDA for exposed names within 12-24 months. Short-term (days–weeks) volatility will track headlines (layoffs, model releases); medium (3–12 months) driven by earnings cadence and capex cycles; long-term (1–5 years) by structural reallocation of labor and corporate capex into AI compute.

Trade implications: Prefer long exposure to high-ROI compute and cloud infra via concentrated long positions (use LEAPs/call spreads 6–12 months) and hedge by shorting legacy silicon/services names that lose market share. Options useful to express skew: buy-call on winners, buy puts on structurally impaired names; expect realized vol to rise 25–60% around major model releases/AI regulation windows.

Contrarian angles: Consensus overestimates near-term job destruction and underestimates demand elasticity—firms may reallocate cost savings into new product lines, supporting capex for winners. Historical parallel: 1990s internet automation reallocated labor and boosted productivity after a 2–3 year disruption; a similar U-shaped recovery is plausible here, so time exposures around 6–18 month inflection points rather than binary long-term bets.

Key Decisions for Investors

• Establish a 2–3% portfolio long in NVDA via 1:1 LEAP calls (9–12 month expiries, ATM to +10% strikes) within 2–4 weeks to capture continued accelerator pricing power; trim if position gains 40% or if bipartisan AI regulation with model/data caps is passed within 6 months.
• Initiate a 1–2% short position in INTC via 6-month OTM puts (10–15% OTM) or a small outright short if margin account allows, scaling in on any quarter-over-quarter revenue misses; target exit on a 30% rally or coverage if management announces a credible node/strategy turnaround within 9 months.
• Implement a paired relative-value trade: go long GOOGL (class A) via a 3–9 month call spread (sell higher strike to finance) at 1–2% portfolio weight and short IBM stock at equal dollar notional (1–2%) to express cloud/AI monetization vs legacy services erosion; rebalance after each company’s next two earnings reports.
• Rotate 3–5% of portfolio from consumer discretionary into cybersecurity and cloud infra (MSFT, ZS, multi-cloud ETFs) over the next 30–90 days to hedge consumption risk from potential job displacement and capture secular SaaS security spend.
• Monitor specific regulatory catalysts: if the US/EU draft AI safety bill within the next 90 days contains model-size, compute or data-sharing restrictions, reduce NVDA/GOOGL longs by 50% and increase cash; conversely, if legislation is permissive or delayed beyond 12 months, add to AI infra longs up to an additional 2%.