Managers & management
If AI Can Assign Work and Track Progress, Do We Still Need So Many Managers?
4 min readAI, Management, Leadership, Automation, MCP
Key takeaways
- Managers spend a large share of their day opening systems, piecing information together, checking status, and chasing delegated work, and AI can now compile much of that into a daily management brief.
- Connecting internal data sources such as meeting notes, chat, and the recruiting system to AI through MCP is what makes a tool like this quick to build.
- Trained on past meetings, messages, emails, and documents, AI can learn more than a writing style; it picks up how a manager sets priorities and moves work forward.
- If AI can assign and track work, that exposes an old problem: managers have spent so much time keeping operations running that the most important parts of management get too little attention.
- Once AI covers a manager's limits on time and information, the manager's value has to show up in developing people, guiding the team to its goals, and turning strategy into daily work.
Mornings spent opening system after system
Every morning, the first thing I used to do at the office was open one system after another: email, team chat, calendar, meeting notes, the recruiting back office. Only after going through all of them could I see the shape of the day. What did I need to handle myself? What had I already delegated? What was waiting on someone else's reply? What was stuck because my boss hadn't made a call yet? Just working that out took real time.
Recently I used an AI tool to build what I call an HR command center, a single page that pulls this together. It came together quickly largely because my company has been steadily connecting its internal systems through MCP (Model Context Protocol). Think of MCP as a universal power outlet for AI. Meeting notes, the chat platform, and the recruiting system were already connected, and email, calendar, and cloud files have ready-made connectors. Once everything is plugged in, the AI can reach the data through the right tool.
Four buckets that describe a manager's day
The command center sorts each day's items into four buckets: "I do," "I'm waiting on," "waiting on my boss," and "FYI." I chose these on purpose. Much of what a manager handles isn't work they complete personally. It's delegating, coordinating, waiting for replies, and pushing other people to decide.
So the tool tracks more than my own to-dos. For delegated work, it records how many business days I've been waiting. After three days without a reply, it reminds me, and one click drafts a follow-up message. Before a meeting, it pulls the conclusions and open items from the previous session in the same series. Afterward, it splits the content into highlights, decisions, tasks assigned to me, and items waiting on others; anything assigned to me becomes a to-do with one click. It regularly checks the weekly action tracker and flags items that have stalled for several weeks in a row. Recruiting progress, staffing changes, and the day's interviews and meetings all sit on the same screen.
The other interesting part is what I'd call self-distillation. I gave the AI my past meetings, messages, emails, and documents and kept training it to follow how I prioritize, how I speak with colleagues, and how I write. Its draft replies now lead with the conclusion. When it writes a delegation note, it covers background, scope, deadline, and expected deliverable in one go. It's learning more than my tone. It's learning how I make judgment calls, get things done, and keep work moving.
Do companies still need managers once AI can assign and track work?
Yes, but a manager's value has to move toward developing people, guiding the team to its goals, and turning strategy into daily work. Building this made me rethink the manager's job. Managers have long spent much of their time gathering information, confirming progress, following up on assignments, and coordinating across departments. Often, simply assembling information scattered across systems eats up most of the day.
Some argue that once AI can assign tasks and track progress on its own, companies won't need as many managers. I see it differently. It reveals a problem with how management has worked: we spend so much time keeping daily operations running that the work that matters most never gets enough time.
If AI can help keep track of progress, organize information, and flag problems, then a manager's value has to show up in whether they can develop people, guide the team to its goals, and turn company strategy into everyday work. Management transformation in the AI era shouldn't stop at how much admin work is removed or how much management cost is saved. The better question is whether, once technology covers a manager's limits on time and information, we can raise the management capability and execution quality of the whole organization.
Three things to take away
- Sort your open items into "I do," "I'm waiting on," "waiting on my boss," and "FYI," and see how much of your time actually goes to waiting and follow-up.
- Set a follow-up threshold for delegated work, such as a nudge after three business days without a reply, instead of relying on memory.
- Decide where the time you save will go, and protect it for developing people, coaching the team, and putting strategy into practice.