AI language models are quickly becoming a key part of knowledge work. They handle many of our administrative tasks, allowing us to focus on the unique work that adds real value.
Microsoft is leading this shift, and many companies now treat AI agents like regular employees, giving them their security, management, and job roles.¹ Experts predict there will be 1.3 billion AI agents by 2028. ²
These AI agents handle routine tasks such as drafting emails, scheduling, entering data, taking meeting notes, and similar tasks.
As AI agents become more common, knowledge workers face an essential question: how will our roles change as AI takes over repetitive administrative work?
Ideally, this change should let us spend more time on deep work, focusing on high-value tasks that use our expertise and creativity.
Still, some of us, myself included, actually enjoy task management and file organization. I sometimes use this time as a break after hours spent diagramming a solution or running a client workshop.
We can only do so much deep work each day—Cal Newport suggests a maximum of 3 to 4 hours.³ So what happens when that’s all we have left? Do we work fewer hours? ⁴ Take a pay cut? Will organizations reduce work hours, or will they expect us to fill the traditional 40-hour work week with more tasks? Could we use this extra time to build new skills or explore new interests?
Knowledge work is changing, for better or worse. We need to adapt, but I’m hopeful we can help shape what this change means for our future.
- ¹ (2025). _Securing and governing the rise of autonomous agents_. https://www.microsoft.com/en-us/security/blog/2025/08/26/securing-and-governing-the-rise-of-autonomous-agents/
- ² IDC Info Snapshot, sponsored by Microsoft, 1.3 Billion AI Agents by 2028, May 2025
- ³ Newport, C. (2016). _Deep work: Rules for focused success in a distracted world_. Grand Central Publishing.
- ⁴ Newport & Cal. (n.d.). _The Four-Hour Work Day_. https://www.thedeeplife.com/podcasts/episodes/ep-259-the-four-hour-work-day