🔮 Does AI make you dumb? And why our forecasts suck #576

2026-05-31 · Source: Exponential View · Field: Business & Management — Corporate Strategy & Leadership, Operations & Process Management, Human Resources & Workforce Development · Depth: Intermediate, long

Summary

A framework by Azeem Azhar and Nathan Warren investigates why individual AI productivity gains often do not translate into proportional firm-level benefits, citing a tech executive's "1+1+1+1=1.5" observation. The analysis critiques equity analysts' linear forecasting models, which systematically underestimate AI's exponential growth, leading to significant revisions in earnings per share estimates for companies like Micron (from ~\$18.25 to ~\$58 for FY2026) and a 40% increase for Google between May 2025 and May 2026. This reliance on "reversion to the mean" risks misjudging AI's impact on hyperscaler investments. Separately, the article explores AI's effect on human cognition, presenting conflicting evidence: college essays show diverse language but less creativity post-ChatGPT, while human Go play improved after AlphaGo. The author argues the key is how technology changes thinking expectations and what incentives foster genuine cognitive engagement, advocating for practices like longhand work alongside AI tools.

Key takeaway

For investors and executives evaluating AI-driven companies, you must recognize that traditional linear financial models systematically misrepresent exponential growth. Your forecasts should account for non-linear adoption and cost curves, avoiding reliance on mean reversion heuristics that lead to significant underestimations. Additionally, for leaders implementing AI, prioritize fostering deep cognitive engagement within your teams by encouraging deliberate, non-AI-assisted thinking practices to prevent skill atrophy and maximize genuine innovation.

Key insights

AI's exponential growth challenges traditional linear forecasting and demands intentional strategies to preserve human cognitive skills.

Principles

• Linear forecasting fails for exponential technologies.
• Cognitive skills can atrophy if offloaded entirely to AI.
• Incentives shape how AI is integrated into work.

Method

The author uses AI for "computer stuff" like transcription and literature reviews, while dedicating two hours weekly to longhand work with pen and paper to foster critical thinking and challenge AI-generated arguments.

In practice

• Dedicate time to longhand thinking without computers.
• Use AI for "computer stuff" and as a critical sparring partner.
• Re-evaluate incentives for deep cognitive engagement.

Topics

• AI Adoption
• Financial Forecasting
• Cognitive Impact of AI
• Exponential Growth
• Shareholder Value
• Human-AI Collaboration

Best for: Executive, Investor, General Interest

Related on AIssential

• How to Think About AI Like a Long-Term Investor — Compound interest applies universally, transforming consistent effort into exponential long-term gains across all life domains.
• In a sea of hype, here are the AI ‘nothingburgers’ you don’t hear about — The AI industry frequently overhypes capabilities, leading to "nothingburgers" and "slop" that fail to deliver real-world value.
• The Future Is Shrouded in an AI Fog — AI's rapid evolution obscures future visibility, challenging traditional long-term investment criteria for leaders.
• Cognitive Synthesis and Neural Athletes — Navigating AI's probabilistic systems requires leaders to unlearn old logic, embrace vulnerability, and become "neural athletes."
• 🔮 Exponential View #560: The $1 trillion panic; my favorite AI analysis tool; intention economy, CAR-T therapy & time++ — Market valuations for AI companies are overreacting due to a fundamental misunderstanding of exponential technology economics and agent demand.
• Why AI Isn’t Transforming Finance Yet — AI adoption in finance is stalled by leadership's traditional mindset and operational practices, not merely technical issues.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Exponential View.