Spending 6.4 Hours a Week on ‘AI Babysitting’! The Hidden Labor of Botsitting is Pressuring Employees

News Overview

• A recent report from the Glean Work AI Institute reveals that white-collar workers are spending an average of 6.4 hours a week on overseeing and correcting AI, a phenomenon dubbed “botsitting.”
• Among the 6,000 workers surveyed, 75% stated that AI has improved their productivity, yet only a mere 13% reported a significant enhancement in overall organizational performance.
• Shockingly, those workers dedicating substantial time to AI “babysitting” are 73% more likely to engage in job-seeking activities compared to their less burdened counterparts.

Key Takeaways

• Uncovering Hidden Labor: Tasks such as inputting context, checking outputs, debugging errors, and correcting mistakes are effectively consuming an entire workday each week to make AI useful.
• The Productivity Paradox: While individual work speeds may increase, employees are bogged down bridging multiple AI tools and completing the contextual work that the AI should be handling, failing to translate into overall corporate benefits.
• Loss of Job Satisfaction: Automation has taken over tasks like relationship building with clients that humans used to enjoy, leaving them with the “boring and unappreciated” job of monitoring AI agents.

Shark’s Eye View (Curator’s Perspective)

Instead of kicking back and letting AI do the heavy lifting, we find ourselves stuck as the “dedicated babysitters” cleaning up the mess AI has made. This is the reality of the workplace in 2026, folks!

What’s particularly noteworthy is the clarity around the term “botsitting.” It’s not just that we’re using AI more; it’s that humans are stepping in to compensate for the parts where the models don’t perform as expected. Organizations ignoring this “invisible labor” and merely competing on the number of AIs deployed will see their top talent gradually swim away back to the job market (the ocean, if you will)!

Despite the claims that “AI agents operate autonomously,” the reality is that humans are desperately tugging at the reins, stripping away the most exciting aspect—creative decision-making. This gap between technological advancement and organizational adaptation is glaring. To bridge it, the answer isn’t “more AI,” but establishing criteria for humans to decide “what not to let AI handle!”

What’s Next?

The phase of just throwing AI at problems is over. Companies that focus on redefining the division of labor between AI and humans through “workflow reconstruction” will be the ones to truly enjoy productivity gains. Meanwhile, companies that impose aimless botsitting will pay the price with soaring turnover rates and declining motivation.

A Shark’s Takeaway

Ending the day as an AI’s babysitter? Not something this shark can watch in silence! The juicy catch (creative work) should be snagged by humans, not left to the bots! 🦈🔥

Glossary

• Botsitting: The ancillary labor of verifying errors in AI-generated content and providing missing context to ensure AI outputs are at a usable level.

• Productivity Paradox: The phenomenon where significant investments in IT and AI fail to yield corresponding increases in statistical productivity, indicating that individual efficiency isn’t translating into organizational success.

• Context Feeding: The act of humans inputting background information or specific business knowledge to enable AI to provide appropriate responses. Inadequate context leads to erroneous AI outputs, making this a crucial aspect of “care.”

• Source: Workers are spending over 6 hours a week botsitting AI, fueling job frustration