Depending on someone’s career experience, any conversation about IC and management roles blending can be triggering. The term Player/Coach makes most designers cringe. It’s often used by companies with unrealistic expectations—expecting one person to do deep individual work while managing an entire team.
As a result, many design managers swung hard in the opposite direction, avoiding the work entirely and centering their role around people and process. As Airbnb’s Brian Chesky once said: “Manage the work.”
Fortunately, as organizations evolve in the age of AI and smaller, high-leverage teams, the distinction between manager and IC is blending again. It’s no longer “you do this, I do that.” It’s a shared responsibility, flexing based on what the moment needs.
On Lenny’s Podcast with Airtable CEO Howie Liu, How we restructured Airtable’s entire org for AI, the concept of the IC CEO was discussed about how even founders need to get close to the work and make to understand the material of AI.
Beyond traditional line management
The concept of line management originates from the Industrial Revolution, when factories needed to organize large groups of workers producing physical goods on assembly lines. The primary management goal wasn’t creativity or autonomy—it was control and coordination.
Frederick Winslow Taylor’s Scientific Management formalized this in the early 1900s. Managers planned, workers executed. The “line” referred to a literal production line, where efficiency was maximized by breaking work into repeatable, measurable units. Productivity was defined by output per hour.
When modern management thinkers like Peter Drucker and Andy Grove came along, they refined these ideas for the information age. Grove’s High Output Management (1983) remains one of the most influential management books in technology—a personal favorite of mine. His central idea was that a manager’s output equals the output of their team, plus the output of the teams they influence.
It’s a brilliant framework, but one built for an era modeled after supply chains and manufacturing processes. Even in knowledge work, Grove described organizations as production lines: inputs, outputs, feedback loops, and yield. Performance reviews mirrored quality control; 1:1s resembled factory inspections.
This system worked when most work was predictable and sequential. Today, we’re operating in a non-linear environment. AI tools can now handle many of the rote tasks that once required human oversight—tracking progress, checking quality, distributing information. The manager-as-foreman model no longer fits.
The opportunity is to reinterpret Grove’s principles for the post-line era:
Output isn’t limited to throughput; it’s creative leverage.
Feedback isn’t inspection, it’s co-creation.
Management isn’t oversight, it’s orchestration.
Instead of managing people on a line, we now manage the flow of intelligence across humans and machines. The lines between thinking, making, and managing blur because they’re no longer sequential. AI collapses the factory floor into a single shared space of real-time context and execution.
In this world, the most effective leaders are not line managers—they’re system designers. They structure environments where both humans and AI systems can do their best work, blending management with making.
Divergence and convergence of roles
Blending doesn’t mean complete fusion. It means the best people can dial in and out of IC and managerial focus depending on what’s needed. Let’s look at where their responsibilities diverge—and where they converge.
IC and Manager focus
Despite increasing overlap, the two roles still have distinct responsibilities. Unless you’re at a completely flat company like Valve, every team needs some structure—to make decisions, allocate resources, and manage outcomes.
Managers focus on hiring, setting direction, feedback loops, and performance management. This is the craft of management—it requires judgment, discipline, and care. These can’t be automated away, but new toolchains can make them more adaptive.
ICs, on the other hand, focus on driving outcomes through craft. Their responsibility is to push the quality of the work, mentor peers, and lead through making. They go deeper than managers can, and that depth becomes their leverage.
Where time is spent together
Being a manager doesn’t make you a leader. ICs can lead, too—but in practice, the balance of authority is uneven. With AI-native tools, that gap is starting to close.
Strategy and direction
LLMs have changed how context and knowledge flow. ICs now have access to the same strategic inputs managers do. Running deep research with AI gives teams a shared foundation, removing the bottleneck where managers used to be the only conduit for context.
Evaluation and analysis
Both ICs and managers evaluate work, but now feedback loops extend beyond user research or support tickets. Teams can analyze prompt performance, evaluate model outputs, and measure quality in real time alongside product and data partners.
Making and co-designing
Managers must prioritize maker time—not necessarily to ship features, but to understand the material. You can’t direct what you don’t understand.
ICs and managers can now co-create, experiment with prompts, and test ideas together. Discovery cycles compress when both roles engage in making.
Exploring new org shapes
The orgs of the past decade were built for predictable scaling—the blitzscaling era of high growth and delivery efficiency. But today’s markets are unpredictable. AI is accelerating change faster than structure can adapt. If you can’t build an org for predictability, build one that can absorb and thrive in change.
In my own team, 80% of my direct reports are individual contributors. It gives us flexibility to pivot quickly. If every manager tried to reorient their entire org at once, it would create too much churn. Instead, IC leads can rapidly redirect toward new opportunities.
Managers still play a stabilizing role: maintaining continuity in key areas, motivating teams, and scaling what works. The number of direct reports may shift, but the span of control—the ability to guide work effectively—remains essential.
A new way of working
Legacy org structures can hold us back. AI reduces the back-and-forth between ICs and managers, creating a shared context in real-time. It’s both challenging and liberating.
To summarize:
Management isn’t going away. It’s transforming.
The new DR is Direct Relationship, not Direct Report. Org charts should provide structure, but should not get in the way of working together
The best managers are becoming systems designers—creating conditions where humans and AI can collaborate fluidly rather than enforcing hierarchy.
ICs and managers now share access to the same strategic context through AI tools, compressing the distance between direction and execution.
The modern org is shaped around adaptive teams, not static roles—where leadership flexes based on expertise, not title.
The modern org is a collective of makers managing a portfolio of opportunities together. There’s still a Head of Design—a first among equals—responsible for clarity and conviction. In this new era, ICs and managers collaborate dynamically, adjusting their energy allocation based on the moment.
Hyperlinks + notes
Prompting isn’t the future. Creating is.
→ One of my favorite reads recently!Tilly Norwood is a gen AI psyop → S1m0ne, the 2002 film starring Al Pacino, prophesied this!