3 min read

Work Design for the AI Era Series #7: Redesign the Role, Not Just the Task

Automating a task changes the accountabilities, judgments, interfaces, and standards surrounding it. Roles must shift toward the value humans now provide—exception handling, quality ownership, customer trust, or decision support—or employees become inconsistent editors of machine output.
3 min read

Work Design for the AI Era Series #9: Managers Become Work System Designers

AI makes status tracking insufficient because output grows faster, appears finished sooner, and creates new review and exception demands. Managers must design how human and machine work flows, set quality bars, allocate attention, and improve the conditions for sound judgment.
3 min read

Work Design for the AI Era Series #10: The AI Work Design Audit

An AI work design audit reveals whether artificial intelligence actually improved how work happens or just got grafted onto your old systems.
2 min read

From Productivity to Throughput Series #1: The Time-Saved Trap

Time saved within a task is not throughput gained across a system. Measuring complete work streams—including queues, approvals, review, rework, quality, and outcomes—shows whether AI actually moves useful work faster or merely accelerates one local step.
3 min read

From Productivity to Throughput Series #2: Local Productivity Is Not System Throughput

Individual efficiency feels good when tasks take minutes instead of hours, but local productivity does not equal overall system throughput.
2 min read

From Productivity to Throughput Series #4: Find the Bottleneck Before You Automate

Focusing on the easiest demo is the worst way to start an AI program because it ignores your actual bottlenecks.
3 min read

From Productivity to Throughput Series #3: Output Volume Can Lower Throughput

Cheap AI output can flood the review, integration, decision, and validation capacity required to finish work. When production outpaces selection, work in progress rises, judgment becomes scarce, and apparent local gains can reduce accepted output across the system.
2 min read

From Productivity to Throughput Series #6: Quality-Adjusted Throughput

Raw output counts can rise while trust, durability, and customer outcomes deteriorate. Quality-adjusted throughput measures accepted, useful work that survives review and avoids rework, revealing whether AI improves the system or simply produces more artifacts.

Choose your reading rhythm.

Start with a weekly briefing, add daily notes, or hear only when a durable essay or research update is ready.
You've successfully subscribed to Antoine Buteau
You've successfully subscribed to Antoine Buteau
Welcome back! You've successfully signed in.
Press ESC to close.