The Return of Multitasking

In the early 2020s, a consensus formed: multitasking is a disease of modern work.

Studies showed it tanked productivity, fried focus, and left us exhausted by the surface-level churn of it all. We nodded along when researchers told us our brains weren’t wired for it. We embraced deep work, timeboxing, and monotasking. The productivity discourse moved on.

Then AI arrived.

And with it, multitasking is making a dramatic return. But, not as a relic of hustle culture, but as something fundamentally different.

The new version of multitasking is all about orchestrating intelligence. Routing focus. Managing “labor” that operates at speeds and scales we’ve never had access to before.

We’re getting to do what chefs and foremen have always done.

In fact, the office worker is becoming both a chef and a foreman.

Chef Mode

Here is what this looks like: four AI tools open at once, each running a different thread.

ChatGPT is deep in research on a topic that came up in a client meeting. Gemini is pulling together a news brief, prepping you for a call in twenty minutes. Claude is helping shape a narrative you’re drafting. Copilot is crunching numbers, working through an analysis that would have taken hours by hand.

You’re bouncing between them. Fast. You’re routing—dropping prompts, moving to the next thing while the previous answer generates.

It feels frantic, but also strangely energizing. Like you’re conducting multiple conversations at once, and somehow keeping all the threads straight.

This is chef mode.

You’re coordinating ingredients, managing timing, and keeping multiple dishes moving at once. It’s surface-level orchestration—fast, fluid, responsive. The AI handles execution, while you handle routing.

This is the new hustle, elevated by AI but still surface-level at its core. Which is why the second mode matters more.

Foreman Mode

This mode of working is completely different.

You’re not bouncing between topics anymore with chat-based AI bots. Here, you are using (the promised, still slightly mythical, but quickly, really arriving) AI agents.

You’re routing focus across a single project, distributing work to agents that handle depth while you manage the whole.

For example, I am building a small personal app. I have OpenAI’s Codex running a bug hunt and cataloging issues. While, in a different panel, Claude Code is working through the bug list, fixing them one by one. Meanwhile, in a separate app, I’ve got one agent building a new feature while another develops a different component entirely. (Aside for my fellow tech lovers, I am bouncing between Google Antigravity, AWS Kiro, and, for a unique take on agentic coding, Nimbalyst. Sometimes multiple at once.)

In these setups, it is about delegating real work to systems that can go deep without you. They’re handling the strategy and the methodologies. The direction, architecture, planning, and quality acceptance.

This is foreman mode.

You’re on the job site, managing skilled trades. The plumber is running pipe, the electrician is wiring circuits, and the framer is building the structure. Each one knows their craft. Your job is to make sure they’re working in sync, that the work connects, that the whole thing comes together.

Microsoft says this type of work will define the Frontier Firm. That companies will be filled with people managing fleets of AI agents, each handling specialized work that used to require full teams. (I’ve now lived it, and am a huge believer in this vision for the future of work.)

For each of us, this is the shift that redefines what multitasking actually means.

Managing the Modes

For some of us, this is here. 
For the rest of you, this is what’s coming.

We get to operate the way chefs and foremen always have: orchestrating work that moves at different speeds, across different layers, without losing the thread.

The form of multitasking that’s always existed on job sites and in kitchens is now arriving at desks. Great hats, outfits, and shoes (optional).