AI Flattening: Why Your Org Chart is About to Disappear

How AI coordination is flattening organizations and making hierarchies obsolete

Disclaimer: the ideas and opinions here don't represent the position of Lattice. These are my ramblings. So, feel free to be brutal and share your thoughts.

A few weeks ago, I started at Lattice after a long stint of working on start-ups in the AI space. While getting onboarded into the HR world, I began thinking about what the future of work could be, especially in a world where AI seems to be the center of every conversation in the Valley. One of the things that struck me was how polarized people are about AI'sposition in the workplace. It seemed like it's either “AI is replacing our jobs” or “AI is empowering humans.” Neither position felt right to me. Instead, I will share my view on a third option: flattening.

This idea sparked during a bike ride around the UW campus in Seattle. My partner made an interesting observation. She predicted that undergrad programs might evolve from rigid disciplines like neuroscience or engineering to broader inquiry-driven questions like “Why do we age?” or "What economic drivers predicted the fall of Rome?" - more like how PhD programs work today.

This got me thinking: would that same idea apply to business organizations?

The Flattening

Here's my core thesis:

AI coordination will enable organizations to flatten from hierarchical org charts into flattened, networks of specialized expertise nodes connected through AI systems.

Instead of the traditional CEO → VP → Director → IC pyramid, imagine organizations where specialized experts contribute their judgment directly to an AI coordination layer that synthesizes insights, identifies conflicts, and facilitates decision-making across the network.

When a strategic decision needs to be made (for example, entering a new market), instead of information flowing up through management layers and decisions flowing back down, the AI system simultaneously queries the relevant expertise nodes: market research for customer insights, engineering for technical feasibility, finance for resource implications, design for user experience considerations. In addition to conducting research and executing on marketing plans & product launches, the AI can synthesize perspectives, identify areas of agreement and disagreement, and presents options with unprecedented clarity and speed. Here, AI acts as delegation engine, positioning human centered work at fueling decisions and opinions critical to the soul of the company.

Why This Could Work

The Diversity Dividend: Research from MIT, McKinsey, and Harvard consistently shows that cognitive diversity drives better decision-making. Teams with diverse perspectives outperform homogeneous teams by 35-70% on complex problem-solving tasks. Yet, traditional hierarchies limit whose voices get heard. AI coordination can surface insights from any expertise node regardless of seniority, maximizing the diversity of input into decisions.

Coordination Overhead Elimination: Brooks' Law demonstrates that adding people to complex projects often decreases productivity due to communication overhead. AI coordination breaks this law by handling the communication complexity, allowing productivity to scale more linearly with expertise quality rather than hitting diminishing returns.

Information Flow Efficiency: Research shows managers spend 61% of their time just gathering and redistributing information. When AI can maintain context across hundreds of conversations and synthesize complex inputs in real-time, this coordination bottleneck disappears.

Speed and Agility: McKinsey research shows that organizations with faster decision-making grow revenue 2x faster than slower decision-makers. AI can coordinate complex multi-stakeholder decisions in hours rather than weeks.

The Size Paradox

When AI handles coordination, productivity isn't constrained by the number of people you have. It's constrained by the number of unique ideas, opinions, and perspectives your team can generate. This aligns with decades of DEI research showing that cognitive diversity drives better decision-making and innovation.

Recently, Sequoia Capital predicted we'd see the world's first single-member unicorn. While that hasn't happened, we're seeing something almost as remarkable: companies achieving massive scale with a handful of people.

Source: Business Insider

Sources: [LinkedIn] Esteban Arroba Del Castillo

It’s Starting…

In Engineering Teams: Instead of a traditional structure where engineers spend 37-44% of their time in meetings coordinating work, AI handles sprint planning, dependency tracking, and progress updates. Senior engineers focus entirely on architectural decisions, mentoring junior developers, and solving complex technical problems.

Example: Atlassian's AI Work Breakdown automatically decomposes large epics into actionable issues in minutes rather than hours, while GitHub Copilot Code Review provides context-aware comments on pull requests and Copilot Autofix automatically suggests fixes for security alerts and ESLint warnings. Result: Engineers can focus on high-value design work instead of coordination overhead.

In Product Management: Rather than PMs spending their days collecting stakeholder input through endless meetings, AI continuously monitors user feedback, market signals, and technical constraints. PMs spend their time on strategic vision, difficult trade-offs, and building conviction around uncertain bets.

Example: Monterey AI integration with Linear automatically triages support tickets, surveys, and call transcripts into Linear issues, while Kraftful AI transforms customer calls and reviews into well-defined user stories with acceptance criteria. This saves product teams hundreds of hours per month on feedback aggregation, allowing them to focus on product direction rather than data collection.

In Customer Success: AI handles routine support tickets, escalation routing, and basic account health monitoring. Human customer success managers focus on strategic relationship building, complex problem-solving, and helping customers achieve transformational outcomes.

Example: Intercom's Fin AI Agent resolves up to 86% of routine support volume, with real-world deployments like RB2B showing 65% deflection rates and saving 132+ support hours per month. Academic studies show that delegating repetitive tasks to AI improves human task performance and satisfaction by 14-15%, as human agents report higher job satisfaction when they only work on cases requiring genuine human judgment and creativity.

Addressing the Hard Questions

"But who's accountable?"

The Surface Answer: AI systems maintain perfect records of who contributed what insights to which decisions. Individuals become directly accountable for their specific expertise and contributions rather than hiding behind management layers.

The Deeper Reality: This fundamentally reimagines accountability itself. In hierarchical systems, accountability flows upward - managers are responsible for their team's outputs, even when they lack domain expertise. In AI-coordinated networks, accountability becomes distributed and expertise-specific.

Consider a product launch failure in a traditional org: the VP of Product takes responsibility, even if the failure stemmed from engineering constraints they never fully understood. In a flattened system, the security engineer who flagged an unaddressed vulnerability, the UX researcher who identified user confusion, and the market analyst who predicted competitive threats all have direct, traceable accountability for their specific contributions.

"What about people development?"

The Surface Answer: Dedicated coaching roles, peer mentorship programs, or AI-assisted skill development replace traditional management development.

The Deeper Reality: We're dismantling the primary career advancement mechanism that has existed for decades. Traditional management ladders provide not just skill development, but identity, status, and financial progression. Removing these requires rebuilding the entire concept of professional growth.

New Development Models Needed:

• Expertise Depth Tracks: Instead of managing more people, advancement means becoming a more valuable expert in your domain

• Cross-Pollination Rotations: Regular exposure to different expertise nodes to build systems thinking

• AI Collaboration Skills: Learning to work effectively with AI coordination systems becomes a meta-skill

• Judgment Calibration: Since human value shifts to judgment and creativity, organizations need ways to develop and assess these inherently subjective capabilities

Identity Crisis: Many professionals derive significant meaning from "managing people" or "leading teams." Flattened organizations need to provide alternative sources of professional identity and fulfillment. This might mean:

• Public recognition systems for expertise contributions

• Teaching and mentorship roles that carry high status

• Rotation through high-impact decision councils

• Ownership of strategic initiatives that span multiple expertise domains

Financial Implications: How do you create compensation growth when traditional promotion paths disappear? Options include expertise premiums, impact-based bonuses, or equity participation that scales with value creation rather than hierarchical position.

"Who makes the final call?"

The Surface Answer: AI for routine operational decisions based on clear criteria. Human judgment for strategic decisions with major uncertainty, but from rotating decision councils or specialized strategy roles rather than permanent hierarchical positions.

The Deeper Reality: This requires developing sophisticated decision classification systems and governance mechanisms that don't exist yet.

The Governance Challenge: Without permanent decision-makers, organizations need dynamic authority structures:

• Context-Sensitive Leadership: Different people lead different decisions based on expertise relevance

• Escalation Algorithms: Clear triggers for when AI should elevate decisions to humans

• Conflict Resolution Mechanisms: What happens when expert opinions fundamentally disagree?

• Democratic Input Systems: Ways for the entire organization to influence strategic direction

The Speed vs. Deliberation Tension: AI coordination promises faster decisions, but some decisions benefit from slow, deliberative processes. Organizations need frameworks for identifying which decisions require speed versus careful consideration.

"What prevents techno-authoritarianism?"

The Surface Answer: Transparency: open algorithms, clear decision criteria, and democratic input into how AI systems operate. Kind of like Wikipedia's governance model, but definitely not Facebook's.

The Deeper Reality: This is perhaps the most critical question.

AI-coordinated organizations could become the most democratic workplaces in history, or the most oppressive.

Transparency Requirements:

• Algorithm Auditing: Regular review of how AI systems make coordination decisions

• Decision Traceability: Clear records of why specific synthesis choices were made

• Bias Detection: Continuous monitoring for how AI systems might favor certain types of expertise or perspectives

• Override Mechanisms: Ways for humans to challenge or reverse AI coordination decisions

Democratic Governance Models:

• Constitutional Frameworks: Clear principles governing how AI systems should operate

• Rotating Oversight: Different expertise nodes take turns auditing AI coordination

• Proposal Systems: Ways for anyone to suggest changes to coordination algorithms

• Exit Rights: Clear processes for experts who disagree with AI coordination to escalate concerns

The Power Concentration Risk: Even in flat organizations, someone controls the AI systems. Preventing techno-authoritarianism requires:

• Distributed Control: Multiple stakeholders involved in AI system governance

• Open Source Coordination: Using transparent, auditable AI systems rather than black-box solutions

• Regular Rotation: Preventing any individual or small group from permanently controlling coordination systems

How We'll Know It's Working

Revenue per employee ratios climbing dramatically faster than inflation

What this looks like: Organizations seeing 2-3x faster revenue growth per employee compared to traditional industry benchmarks, with profit margins expanding as coordination costs plummet.

Some Leading Indicators:

• Reduction in coordination overhead (30-50% fewer status meetings)

• Increase in deep work time per employee (60%+ time on domain expertise)

• Faster decision cycles without quality degradation

• Higher expertise utilization rates across all domains

Critical to watch: Short-term productivity gains that don't sustain, quality degradation if AI coordination misses nuances, employee burnout from unrealistic pace expectations.

Decision speed and quality both improving

The breakthrough: Traditional wisdom suggests speed and quality are inversely related. AI coordination breaks this by surfacing relevant expertise faster, identifying potential issues before implementation, and maintaining context across multiple simultaneous decisions.

Success markers:

• Time from problem identification to solution implementation (50%+ faster)

• More stakeholders involved without slowing decisions

• Reduction in decision reversals due to incomplete information

• Innovative solutions emerging from diverse expert synthesis

Job satisfaction rising through expertise focus

The transformation: People spend 60-80% of their time on pure domain expertise rather than coordination overhead, leading to higher engagement and retention of top performers.

Key changes:

• Stress shifting from organizational politics to domain challenges

• Recognition based on expertise impact rather than hierarchical position

• Autonomous decision-making within areas of expertise

• Cross-pollination learning between different domains

The Pattern: AI handles coordination, humans provide expertise and judgment. Organizations become networks of specialized knowledge rather than pyramids of authority.

The transition to AI-coordinated flat organizations represents one of the most significant shifts in how humans organize collective work since the industrial revolution. Success won't just be measured in efficiency gains, but in whether these new structures enhance human agency, creativity, and fulfillment while delivering superior organizational outcomes.

The Hard Truth

This transition will be disruptive. Many current management roles will become obsolete, not because AI replaces managers, but because the coordination functions they serve become unnecessary. The organizations that figure this out first will have massive competitive advantages.

However, it's not inevitable. Organizations have to actively choose to restructure around AI coordination rather than just using AI to make existing hierarchies more efficient. Most will choose the safer path of incremental improvement.

Start-ups, on the other hand, must evolve mechanisms to compete with well-placed incumbents. They may rely more heavily on flattening to scale their business or maintain a small size with heavy technology use to remain nimble. Sufficed to say, we are seeing the next generation of organizational design emerge from highly productive start-ups.

The future of work isn't about humans versus AI or even humans plus AI. It's about humans and AI creating entirely new organizational forms that were impossible before now.

This requires being honest about what humans are uniquely good at, investing in developing those capabilities and designing organizations that prioritize human agency rather than algorithmic efficiency.

The question isn't whether AI will change how we work. It already has. The question is whether we'll use it to create more human workplaces or less human ones.

That choice is still ours to make…