Using AI at work can lead to ‘brain fry’

A Boston Consulting Group study of nearly 1,500 U.S. full-time workers finds AI oversight raises mental effort by ~14% and information overload by ~19%, producing what researchers call “brain fry.” Productivity gains are non-linear: using 1→2 AI tools gives a meaningful boost, a third tool adds less, but >3 tools reduces productivity. While AI that replaces repetitive tasks lowered burnout (physical/emotional), increased oversight and workload raised cognitive fatigue, error risk and intention to quit—implications for workforce productivity and turnover.

Analysis

AI’s real friction point is oversight and tool proliferation, not model quality — that means the first wave of demand will flow into orchestration, explainability, logging and human-in-the-loop platforms that reduce cognitive switching costs. Expect CIO budgets to reallocate from 10–30 small point solutions into 1–3 platform buys over 12–24 months; that consolidation compresses multiples for standalone niche vendors while expanding TAM for integrators and cloud-native governance.

A second-order labor effect is faster churn in high-touch functions (marketing, ops, product) that will force incremental spend on HR analytics, employee experience tooling and managed services; I would pencil a 5–10% incremental line item for these categories in enterprise IT budgets over the next 12 months. Regulatory and audit pressures—recordkeeping of model decisions and demonstrable human oversight—are plausible catalysts that could accelerate spend on compliance-oriented tooling within 6–18 months.

The equilibrium outcome is platform concentration: winners will be those that stitch telemetry, identity, explainability and workflow into a single pane; losers are small, single-feature AI apps that increase oversight burden. That outcome is not priced in for many mid-cap AI app vendors today and creates a near-term trade window around names exposed to fragmentation vs. orchestrators who can capture pricing power as firms simplify their stacks.

Key Decisions for Investors

• Long Accenture (ACN) 6–12 months — buy a 9–12 month call spread (moderate debit) to express managed-services and integration tailwinds. Rationale: ACN is positioned to sell migration + governance packages; reward if 5–10% incremental spend materializes, risk is macro-driven IT cutbacks. Target return 1.5x–3x if consolidation accelerates.
• Long UiPath (PATH) 6–12 months — buy PATH outright or a 9–12 month call (or call spread) to play automation-as-remedy for repetitive-task burnout. Rationale: automation reduces burnout and HR churn, creating stickier ARR; downside: slower enterprise rollout or execution missteps. Risk/reward: asymmetric: limited downside vs >2x upside if enterprise adoption lifts ARR growth.
• Long Snowflake (SNOW) 9–18 months — buy LEAPS or 12 month ITM calls to capture centralization of outputs from multiple AI tools into a single data governance layer. Rationale: SNOW sits at the aggregation layer where orchestration reduces cognitive load; reward if customers consolidate tooling and increase consumption, risk is competition from cloud hyperscalers. Expect 1.5x–3x upside in a consolidation scenario.
• Short C3.ai (AI) 3–12 months — single-stock short or buy put spread to hedge exposure to over-indexed pure-play AI vendors that lack governance/integration offerings. Rationale: market may mark down niche AI app multiples as buyers favor integrated platforms; risk is positive sentiment from new contracts or infra tailwinds. Aim for 20–40% downside capture in a 6–12 month window.