New Management: From Doing the Work to Directing AI Agents
4/24/25
What if everything we "know" about individual contribution is about to become obsolete?
The notion that career advancement comes from becoming better at doing specialized work yourself is colliding with a new reality: AI agents are taking over execution while humans are stepping into orchestration. We're still building career ladders that reward personal output while the market increasingly values the ability to direct intelligent systems.
From Solo Performer to Orchestra Conductor
30%
Organizations
Will have digital workers as core contributors by 2025
85%
Enterprise Interactions
Will be managed without human involvement
92%
Companies
Are investing in AI capabilities
1%
Organizations
Believe they've reached AI maturity
Look, the data doesn't lie here. By 2025, digital workers will be core contributors at 30% of organizations, with 85% of enterprise interactions managed without human involvement. We're not just talking about chatbots answering basic questions – we're witnessing the emergence of sophisticated AI agents capable of performing specialized knowledge work that once required years of human expertise.
I've spent the last three months interviewing technical specialists across industries who've found themselves unexpectedly managing teams of AI agents rather than doing work themselves. A common refrain emerges: "I was hired to do the work, but now I spend most of my day directing AI to do it instead."
This isn't automation stealing jobs. It's a fundamental transformation in how value is created. The skills that traditionally defined career success – technical depth, execution speed, attention to detail – are being augmented by completely different capabilities: instructing AI effectively, designing workflows across multiple agents, validating outputs, and continuous system improvement.
Consider this paradox: while 92% of companies are investing in AI capabilities, only 1% believe they've reached AI maturity. This isn't just a technology gap – it's a profound misalignment between how organizations develop talent and where work is actually heading.
The Invisible Career Transition
The transition from doer to director isn't showing up in job descriptions yet, but it's happening under the surface across every knowledge work domain. A software engineer I interviewed recently confessed: "I write maybe 10% of the code I used to – the rest is prompt engineering, reviewing AI-generated code, and designing how multiple AI systems should interact."
The New Reality
Microsoft's ecosystem already features AI agents that handle everything from project management to reconciling financial statements. As Jared Spataro, Microsoft's Chief Marketing Officer for AI at Work puts it: "Think of agents as the new apps for an AI-powered world." But unlike traditional apps, these agents learn, adapt, and increasingly work with minimal oversight.
This creates a strange contradiction in how we evaluate performance. We're still measuring individual contributors by their personal output while their actual value increasingly comes from how effectively they direct AI systems to produce that output.
The Unbundling of Expertise
Imagine a radical new vision for technical career paths - one where high-performers spend 80% of their time managing AI instead of doing specialized work themselves.
The traditional "expert" is being transformed into two distinct roles:
1
1
The Knowledge Itself
The knowledge itself, increasingly embedded in AI systems
2
2
The Orchestrator
The orchestrator who knows how to direct those systems effectively
As BCG research shows, supervising virtual AI agents is becoming a critical "teaming skill" for organizations deploying intelligent systems. But our performance reviews and job descriptions haven't kept up with this reality.
Consider a financial analyst who previously spent days building complex models. Now, she accomplishes the same work in hours by directing multiple AI agents - one for data extraction, another for modeling, a third for scenarios, and a fourth for visualizations. Her value lies in orchestrating these systems to produce a cohesive, validated output.
The New Management Hierarchy: Humans at the Top, AI in the Middle
Fading Fast
- Executing specialized tasks efficiently
- Memorizing domain knowledge
- Consistent personal productivity
Rising in Value
- Effectively instructing and guiding AI systems
- Designing workflows across multiple agents
- Critical assessment of AI-generated outputs
- Pattern recognition across systems
The irony is striking. After decades of "flattening" organizational hierarchies, we're witnessing the emergence of a new management layer – not of human middle managers, but of AI agents requiring human direction.
McKinsey senior partner Jorge Amar notes that "Companies are going to need to rewire how functions work so they can get the full value from gen AI agents." This rewiring isn't just about technology implementation – it's about fundamentally reconsidering what we value in individual contributors.
The capabilities that differentiate top performers are rapidly shifting.
From Linear to Exponential Impact
Traditional Contribution
One person creates linear impact
AI Integration
Humans direct intelligent systems
Exponential Results
Multiplied impact through orchestration
The most fascinating aspect of this shift is how it transforms the productivity equation. Traditional individual contribution creates linear impact – one person can only do so much, no matter how talented. But an effective AI agent manager potentially creates exponential impact.
PwC describes this evolution as "human-led and agent-powered," where employees "refocus on doing what only people can – guiding and overseeing AI, innovating with AI's help, and making swift, AI-supported decisions."
I recently spoke with a marketing strategist who previously managed campaigns personally but now orchestrates a team of specialized AI agents – one analyzing market data, another generating content variations, a third optimizing ad placements, and a fourth analyzing performance. She's effectively directing what would have previously required a team of specialists, multiplying her impact by an order of magnitude.
For individual contributors, this means that career advancement won't come from becoming incrementally better at executing tasks yourself, but from dramatically expanding your ability to direct multiple AI agents toward increasingly complex outcomes.
Navigating the Great Shift
What does this mean for individual contribution?
Beyond all the tactical implications we've explored, there's a more profound transformation happening: the redefinition of what "individual contribution" actually means. The most valuable individual is no longer the virtuoso soloist but the skilled conductor, coordinating an orchestra of intelligent systems to create something far beyond what any individual could produce alone.
How will organizations need to adapt?
The organizations that thrive won't be those with technically brilliant people who insist on doing everything themselves, but those whose people excel at delegating to and directing AI agents while focusing their human capabilities on what machines still can't do well.
What should you be considering about your own role?
Consider your own role – what percentage of your tasks could potentially be performed by AI agents under your direction? What would your impact look like if you spent most of your day orchestrating AI systems rather than performing tasks yourself? Are your professional development efforts focused on becoming a better doer or a better director?
The future doesn't belong to those who can outperform AI at specialized tasks. It belongs to those who can direct AI to deliver outcomes that neither humans nor machines could achieve alone.