AI Breaks the 50-Year-Old Law That Governed Software Companies: Here’s What Takes Its Place

(SeaPRwire) –   Back in 1975, a software engineer named Fred Brooks released a management book outlining the innate challenges of scaling technology companies. He titled it The Mythical Man-Month, and the name pointed to a straightforward core insight: adding more personnel does not translate to faster project delivery.

To put it simply, ramping up the output of a software team operates entirely differently from increasing production for workers at a physical goods factory. Hire 10 extra factory employees, and you get 10 additional units of product. But invest 10 times more capital and bring on 10 times as many programmers, and you will not end up with 10 times as many functional lines of code.

Brooks derived this conclusion from firsthand experience. While working on IBM’s 360 mainframe operating system project, he witnessed software teams collapse under the weight of their own growing complexity. Each new hire drove communication costs up exponentially. New staff required training, and their onboarding period meant they contributed very little to output at first. Existing team members, meanwhile, had to pause their own work to train the new hires — a double setback that grew worse with every additional employee brought on board.

For 50 years, no one found a viable workaround for this rule. Of the 66 unicorns (startups valued at over $1 billion) that had ample cash reserves in 2021, 30 have not secured any further funding since that year, and 11 have raised new capital at reduced valuations. While other factors almost certainly contributed to these outcomes, this trend serves as additional evidence that productivity cannot be purchased simply by hiring more engineers.

That all began to shift in 2022, however.

Why AI Repeals Brooks’s Law

Starting in 2023, a new set of principles has begun to govern how investment capital is allocated, principles that have effectively rendered the Mythical Man-Month* obsolete. This is easy to observe when looking at companies funneling large sums into AI model development and seeing immediate gains in research progress and model performance. AI model firms have managed to deploy far more capital with much smaller teams, and have generated extraordinary revenue growth as a result. Our internal data, in fact, shows that the largest AI companies have a revenue run rate per full-time employee that is nearly three times higher than that of non-AI software and technology firms.

The cause of this shift goes far beyond improved tools or more efficient workflows. Contemporary AI development methods have evolved to rely on massive volumes of computing power rather than intricate, layered engineering work, which means the longstanding coordination issues caused by team complexity have largely disappeared. Computer scientist Rich Sutton famously laid out this dynamic in his 2019 essay “Bitter Lesson”, in which he argued that simple algorithms running on powerful computing systems consistently outperform sophisticated algorithms built on human domain-specific expertise. When Sutton published the essay in 2019, ChatGPT did not exist, and no one was conducting hundred-million-dollar training runs for advanced AI models. The subsequent rise of cutting-edge frontier AI has validated his argument far more dramatically than almost anyone anticipated at the time.

Brooks’ decades-old observation holds true for traditional software development. AI development, by contrast, follows entirely different rules. Instead of requiring large teams spread across multiple interconnected subsystems that need constant cross-team coordination, AI models are built by much smaller teams whose output quality scales directly with the amount of data and computing power they have access to. The end result is a reality Brooks would have found nearly impossible to imagine: capital can finally be deployed rapidly, and the link between investment and output is far more direct. Put simply, you now can pour money into software engineering to get higher output — that is, if you are building AI models that carry out the tasks we once relied on traditional software to complete.

What the New Numbers Look Like

This shift is already playing out in private markets, where companies are raising record amounts of capital with historically small teams, and posting unprecedented growth rates. OpenAI, Anthropic and Cursor have all grown their revenue from just a few million dollars to billions in less than two years.

The formula for success has also changed. For a long time, the solution to the limits outlined in the Mythical Man-Month* was stronger leadership and more cohesive organizational culture. Better-managed teams outperformed competitors by executing faster and more efficiently even with the same amount of capital. Recently, however, AI has shifted the main bottleneck from human talent to computing power, and the ability to manage large teams at scale is far less of a competitive advantage than it once was.

The limitation Brooks identified was always a supply-side issue: it was impossible to build successful software companies quickly enough to keep up with market demand. This same dynamic extended to the venture capital space: funding was abundant, but high-quality companies capable of putting that funding to good use were not. This pattern has held across multiple market cycles, with investment returns concentrated in a small number of standout firms, and no amount of available funding altering how many of these standout firms emerge in any given period. But this shortage of standout companies was never caused by a lack of ideas or capital. It stemmed directly from the dynamic Brooks observed: you could not scale companies on demand. Change that limitation, and you eliminate that scarcity.

What Comes After The Mythical Man Month
The ramifications of this shift will be far-reaching. Investment gains will flow to those who can deploy capital quickly and effectively, not just to parties with the sharpest consumer insight, strongest work ethic or best leadership skills. For both customers and investors, this could mean far more opportunities to build generation-defining companies without facing fundamental scaling limits.

The pace of change across the entire software industry, and every sector it impacts, will only keep accelerating. The scope of areas that software can transform will also expand. Before the rise of AI, engineering progress was limited by the diminishing returns that came with adding more programmers to a project. With AI, the global tech ecosystem has found a way to work around that longstanding barrier.

Brooks identified the bottleneck that the software industry navigated for more than 50 years. The limitations laid out in The Mythical Man Month were long considered impossible to overcome. But in the AI era, success may come down to nothing more than a sufficiently large computing budget, paired with a small, skilled team that knows how and when to leverage that computing power.

The views shared in .com commentary pieces are exclusively those of their authors and do not necessarily represent the official stances or beliefs of .