Skip to content

Podcast

Botsitting- The Work Draining AI Gains

The AI Daily Brief: Artificial Intelligence News and Analysis

Source ↗ ← All highlights
  • Botsitting Is Hidden Work Consuming AI Gains
    • Botsitting is a new, largely invisible form of labor where workers feed context, check outputs, debug mistakes, and clean up AI’s confident but wrong answers.
    • The report finds workers spend 6.4 hours/week botsitting, which consumes much of the average 11 hours/week saved by AI. (Time 0:02:16)
  • Botsitting Occupies The Largest Slice Of AI Time
    • AI time splits into building/learning (27%), active use (36%), and botsitting (37%), making botsitting the single largest category of AI time.
    • Within botsitting, feeding context (2.3h), supervising outputs (2.2h), and debugging (1.7h) are the main drains. (Time 0:05:07)
  • Feeding Context Rapidly Raises Burnout Risk
    • Botsitting causes an exhaustion multiplier: every 10% more time feeding context makes workers 25% more likely to feel worn out.
    • Frequent botsitters (40%+ of AI time) are 73% more likely to be actively job hunting. (Time 0:06:30)
  • Botsitting Encourages Offloading Responsibility
    • Botsitting can cause moral disengagement where people increasingly blame AI for mistakes and offload responsibility.
    • The report finds heavy AI users are 3.4x more likely than light users to blame the tool when things go wrong. (Time 0:07:44)
  • Smarter Tools Can Make Workers Sloppier
    • Agentic AI amplifies botsitting because smarter tools increase temptation to stop verifying outputs and users may lack the capability to check novel outputs.
    • Nathaniel argues this dynamic likely grows as agentic adoption rises, not shrinks. (Time 0:13:31)
  • Trying New Models Shows Verification Limits For Nonexperts
    • Nathaniel shares his experience testing new models like Claude Code and feeling limited to impressionistic judgments because he doesn’t code.
    • This illustrates democratization of skills but highlights verification challenges when users can’t validate AI outputs. (Time 0:13:45)
  • Build Human Infrastructure Around AI
    • Build a human infrastructure for AI at individual, team, and organizational levels to turn individual gains into systemic transformation.
    • Focus on how people work with AI, how teams manage it, and how organizations design processes and governance around it. (Time 0:15:41)
  • High Achievers Reinvest AI Time Into Skills
    • High AI achievers spend less AI time on core tasks and more on learning, reinvesting saved hours into skill-building and using botsitting productively.
    • They treat AI as a teacher and deliberately orient botsitting toward improving outputs and expertise. (Time 0:15:59)
  • Use AI To Cut Coordination, Not Replace Management
    • On teams, keep human accountability even while treating AI as a teammate and use AI to remove coordination costs.
    • High-achieving teams delegate coordination to AI and reinvest manager time in people instead of replacing management. (Time 0:18:31)
  • Transformative Organizations Make AI Governance Living
    • Transformative organizations measure relevant metrics, give employees visibility into AI usage, treat governance as living, and invest heavily in people and training.
    • Examples: 71% of employees at transformative firms can see their AI usage vs 40% elsewhere; 93% review AI policy regularly. (Time 0:21:36)