Automation Mayhem: An AI Revolution Update

June 22, 2025

Price's Law: When 10% of People Do 50% of the Work (Now on AI Steroids)

In any large team, a few superstars carry much of the load. Price's Law states that roughly the square root of the number of people produces half the output. In a project of 100, about 10 overachievers generate 50% of the results. This imbalance is being supercharged by AI. Top performers armed with AI tools can become even more productive, leading to an extreme concentration of productivity. The "vital few" can leverage AI to accomplish work that used to require entire departments, effectively making the heavy tail of productivity heavier.

This is not theoretical. AI startups are reaching $100M in annual revenue with under 100 employees, an unprecedented scale of efficiency. For example, Lovable hit $17 million ARR in 3 months with just 15 people (over $1M revenue per head!). Midjourney reportedly generates $4.1 million per employee, and even OpenAI is estimated at $1.5M revenue per worker. These ratios show that with AI, reaching $5M, $10M, or even $100M in revenue with only a handful of humans is now plausible.

AI amplifies the creative and productive force of top talent, enabling unprecedented economies of scale. A single brilliant coder with AI assistance can potentially replace dozens of average coders. This "winner-take-most" dynamic, already visible in software, is further accelerated by AI. We may be approaching the first one-person, billion-dollar company. The implications for inequality within organizations are profound: a small number of AI-empowered employees might contribute the bulk of value, leaving others to find new niches or face redundancy. It's Price's Law on steroids: in the AI era, the rich (in productivity) get richer.

On a broader scale, AI contributes to industry concentration. A few giant firms dominate markets, mirroring how a few individuals dominate output within companies. AI amplifies this, as those with more data and computing power (usually larger players) can create better AI, leading to a feedback loop of dominance. The digital revolution already amplified wealth and productivity concentration; AI accelerates this, creating even greater scale advantages for those at the top. The corporate world is becoming like an MMORPG where a few high-level wizards (with AI familiars) do most of the questing.

Outsourcing to Algorithms: Lean Teams and the New Efficiency

Companies have historically outsourced work to countries with cheaper labor. Now, businesses can outsource tasks to machines. AI systems are taking over routine, repeatable tasks, functioning as ultra-efficient outsourcers. This puts downward pressure on hiring: why hire an entry-level employee when a digital worker can do the job instantly and at near-zero incremental cost?

We are already seeing this shift. AI excels at automating repetitive tasks, like using chatbots for customer service or automated data-processing tools for back-office functions. This leads to dramatic boosts in speed and cost savings. Even traditional outsourcing firms are pressured to integrate AI to remain competitive.

This trend creates leaner teams. Companies can maintain a small headcount because much of the "heavy lifting" is done by software. This is a revival of the Lean Startup ethos, or "Lean Startup 2.0." A tiny startup can achieve outsized productivity, potentially reaching massive revenues on just a seed round without needing large teams. The need to hire armies of people is reduced when each employee is armed with AI copilots and cloud services.

This shift also highlights the importance of human limits like Dunbar's Number. Robin Dunbar's research suggests we can only maintain stable social relationships with roughly 150 people. Historically, companies grew large despite communication breakdowns because many humans were needed. If AI allows big things to be done with fewer humans, businesses may choose to stay small and cohesive, leveraging technology for scale instead of headcount. Smaller teams under the ~150 threshold excel in agility and unity, while larger ones see cohesion erode and bureaucracy creep in. Savvy companies like Gore-Tex famously capped plants at ~150 people to preserve a tight-knit culture. In the future, more firms might intentionally keep divisions human-sized, letting software handle the rest. Why deal with complex org charts if a tight crew of 50 or 100, equipped with AI, can outperform a bloated staff of 1,000?

In short, AI-enabled outsourcing to algorithms means doing more with fewer people. It's economically efficient, but it raises societal questions. Outsourcing in the 2000s shifted jobs across borders; in the 2020s, it shifts jobs to machines. Companies will need to navigate the human impact: re-skilling staff for higher-value roles and addressing morale when AI takes over parts of jobs. The optimistic view is that AI frees employees for creative tasks; the cynical view is that many doing routine work will be laid off. Either way, the lean, AI-heavy team is becoming a competitive staple.

The Monkeysphere at Work: Dunbar's Number and Organizational Cohesion

Dunbar's Number (~150) is the "monkeysphere" of a workplace – the limit beyond which colleagues become "that random guy from accounting" rather than individuals. Human brains evolved to trust and cooperate in groups up to a certain size; beyond that, informal trust gives way to formal rules and bureaucracy.

AI might enable new company structures that respect these cognitive limits. Instead of large monolithic corporations, we could see networks of smaller, tight-knit teams (roughly Dunbar-sized or less) loosely coupled through technology. Think "Hollywood model" organizations: small expert crews coming together for a project, powered by AI tools, and disbanding or re-forming as needed. When software handles coordination and grunt work, massive permanent hierarchies are not needed for complex products or services. Smaller teams can achieve more on their own and collaborate via platforms without merging into one giant organization.

This benefits agility and cohesion. Studies show that beyond ~150 people, communication silos form and engagement drops. Staying small helps companies retain a shared culture and purpose. There's less need for complex org charts and middle-management fiefdoms. Companies like W.L. Gore deliberately structure around Dunbar's Number to foster innovation.

AI could make such models even more powerful. Small teams enhanced with AI can punch far above their weight. A "Team of 5" might manage what used to be a 50-person operation, with AI assistants handling customer support, data analysis, and marketing. The humans strategize and build relationships. These five can still easily collaborate, ideally aligning with Jeff Bezos's "two-pizza team" rule – big enough for diverse skills, small enough to stay cohesive, and amplified by AI.

Large enterprises might reorganize into many small, semi-independent teams, leveraging common AI platforms. Internally, this would feel like a swarm of startups rather than a top-down bureaucracy. This structure leverages human social strengths (tight trust circles) while using AI for coordination at scale. It's scaling without actually "scaling" headcount beyond diminishing returns. In an AI-rich environment, bigger isn't better for headcount; many small, AI-empowered pods are preferable to one bloated behemoth. People aren't wired to meaningfully connect with hundreds of coworkers. Keep the tribe small, and let the machines bridge the gaps between tribes.

Automation at Every Level: From Interns to CEOs

When we discuss AI taking jobs, we often think of factory or clerical roles. However, automation pressure is affecting both the bottom and top of the organizational chart. Entry-level and routine jobs are directly impacted: AI can draft emails, schedule appointments, crunch numbers, write code, or design basic graphics, threatening roles like administrative assistants, junior analysts, and customer service reps. By 2030, about 30% of US jobs could be fully automated, and 60% of occupations significantly altered by AI tools. Many traditional "junior" roles risk being eliminated or drastically transformed; your next administrative assistant might be a bot.

Senior leadership and management roles are also affected, albeit differently. While robo-CEOs aren't imminent, AI is encroaching on tasks high-level managers perform. AI can analyze vast datasets for strategic insights, run simulations, and provide real-time recommendations to executives. AI advisors can suggest actions based on analytics. This means AI is taking on some of the decision-making burden previously exclusive to upper management.

Furthermore, AI can flatten hierarchies by enabling self-management. If an AI system can coordinate projects—assign tasks, monitor progress, and nudge people—fewer project managers might be needed. AI-based project management tools are already optimizing workloads and predicting timelines in software engineering. For routine supervisory tasks like expense reports or scheduling, AI can automate administrative overhead, potentially reducing the need for many middle managers whose jobs involve paperwork and pipeline management.

For the C-suite, AI can provide comprehensive analyses, sometimes challenging leaders' gut instincts. An AI might recommend a cost-cutting measure based purely on efficiency, lacking human empathy. Leaders might override the AI, but they now have to consider "what the algorithm thinks." In some cases, boards might question a CEO who doesn't follow AI recommendations. A highly AI-driven company could potentially replace a poor-performing executive with someone acting as a human interface for AI's strategic plans.

However, human leadership is not obsolete. People offer soft skills, creativity, and moral judgment that AI still lacks. The consensus is AI-augmented leadership, not AI replacing leadership. Future managers might offload number-crunching and rote supervisory tasks to their AI counterparts.

Conclusion

In sum, the AI revolution isn’t just giving us smarter machines... it’s pushing us to reimagine the size, shape, and dynamics of organizations. The companies of tomorrow might be unrecognizable: some as small as a pizza party, some as huge as a planetary network of algorithms, and many in between that won’t last long. The workforce of tomorrow will blend human and AI skills so intimately that “colleague” might mean the person in the next chair or the AI system on your screen. And the most successful organizations will figure out how to make all these pieces – the brilliant 10%, the cohesive 150-person unit, the tireless AI assistants, the accountable humans, and the supportive platforms – work together in a new harmony. It’s a wild, fascinating time to be part of the labor force (or what’s left of it). Just remember: bring your brain, bring your creativity, and don’t forget to bring your AI – you’ll need all three.

Sources