Ten Reasons Why We Won’t See Productivity Improvements from GenAI

2026-02-20 · Source: Tom’s Substack · Field: Business & Management — Corporate Strategy & Leadership, Operations & Process Management, Economic Analysis & Policy · Depth: Intermediate, medium

Summary

An analysis outlines ten reasons why generative AI (GenAI) is unlikely to deliver substantial individual productivity improvements or significant business value. Challenges include the difficulty of end-to-end process redesign, inadequate user training, and the fact that "right way" GenAI use (prompting, review, editing) often negates time savings. Measuring aggregate individual productivity gains is complex, and how employees utilize saved time is unclear. Rising token costs erode labor savings. Most individuals do not invest in personal AI agent development, and organizations rarely conduct controlled experiments. Poor GenAI outputs, termed "workslop" and "process slop," actively decrease overall productivity and trust. Even advanced AI agents require monitoring, limiting benefits. US nonfarm productivity growth, at 2.1% annually over seven years and 0.3% in Q1 2026, shows no clear GenAI-driven acceleration, suggesting market overvaluation of AI providers.

Key takeaway

For AI/ML Directors evaluating GenAI's impact on organizational productivity, recognize that individual gains are often illusory. You must prioritize end-to-end process redesign and invest in highly segmented, job-specific AI training. Do not rely on generic usage or expect significant time savings from "right way" GenAI use. Implement controlled experiments to measure actual productivity and quality, accounting for rising token costs and the risk of "workslop" degrading downstream efficiency. Your focus should shift from individual task acceleration to systemic transformation.

Key insights

GenAI's individual productivity gains are often offset by process, training, measurement, and quality issues, hindering overall economic impact.

Principles

• Substantial AI value requires end-to-end process redesign.
• Effective AI training must be highly segmented and job-specific.
• Rigorous GenAI use (prompting, editing) often consumes time savings.

In practice

• Conduct controlled experiments to measure GenAI impact.
• Develop highly segmented, job-specific AI training programs.
• Account for token costs when assessing GenAI productivity gains.

Topics

• Generative AI
• Productivity Measurement
• AI Training
• Process Redesign
• AI Agents
• Token Costs
• Workslop

Best for: Executive, AI Product Manager, Product Manager, Director of AI/ML, Consultant, Investor

Related on AIssential

• GenAI Success Metrics: Look Beyond Reduced Workload — GenAI's true value often lies in reshaping workflows and improving qualitative outcomes, not just reducing time.
• AI Weekly Issue #509: AI Productivity: it works best for the people losing their jobs — AI boosts novice productivity on verifiable tasks but can hinder experts, creating a "jagged frontier" of utility.
• Measuring the Self-Reported Impact of Early-2026 AI on Technical Worker Productivity — Self-reported AI productivity gains, though often overestimated, are best measured by "value" and complement other assessment methods.
• The Innovation Advantage GenAI Can’t Give You — Generative AI makes ideas a commodity, shifting innovation advantage to human insight in identifying and reframing core problems.
• Is Generative AI creating more Jobs? — Generative AI's impact on job markets presents a paradox of efficiency-driven layoffs versus potential long-term job creation.
• 🔮 Why AI isn’t showing up on your bottom line — AI's firm-level ROI lags individual productivity due to organizational "congestion," mirroring historical general-purpose technology adoption patterns.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Tom’s Substack.