Article
Symbiotic Firms

The point of a firm is that, under one roof, individuals can accomplish more together than they could on their own. If the firm serves its purpose, the whole is greater than the sum of its parts.
In academic and demo environments, AI agents produce better work at lower cost, with gains that are promised to eventually show up in the economy. But the path from individual AI capabilities to economic outcomes is circuitous because existing firm processes and organizational structures coordinate teams of humans. The gains one might naively expect when dropping agents amidst humans in a firm built for a previous era do not materialize. The micro-AI promises, therefore, do not deliver macro value. Going forward, firms that are not set up to coordinate thousands of AI agents running in parallel may clumsily deploy swarms of agents that, at best, are ineffective, and, at worst, wreak havoc far beyond the firm walls.
Does the productivity promise of The Firm break down with AI agents? It doesn’t have to.
To prove this out, we took a more direct (and safer) path by designing AI-native professional services firms from first principles. An agentic core shapes the business model, team structure, and client service models. Because we build and power firms, rather than merely sell tools to them, we can measure technology and process value directly through quality, volume and cost of output, dramatically shortening the path from technical decisions to outcomes and giving us a tight feedback loop to invest more in what demonstrably drives the most value.
For example, Norm Law is an operating environment that builds, tests, and improves an agentic platform to better serve clients. Practicing attorneys leverage agents as much as they can to the clients’ benefit, while simultaneously ensuring human expert supervision of agent output. This is the best of both worlds: human ownership and control of outcomes, along with maximum deployment of AI agents. In a Symbiotic Firm like this, the people and the AI inextricably depend on, and improve, each other.
AI companies are trying to sell more tokens with forward-deployed teams that map company systems and bring their AI into existing flows. But these are the wrong systems to map and optimize. Pre-agent human coordination mechanisms are largely irrelevant. Instead, we grow firms in an AI-native environment from inception to allow the operating cadence and systems to emerge organically as the firm completes more work to the benefit of clients.
With a front-row seat to experts doing consequential work with machines every day, the platform can learn how people delegate, supply context, judge results, correct mistakes, and even evolve their approaches as the machines become more capable through this build. The integration between people, technology, and operations, built from the ground up to thrive amid fast-increasing AI capabilities, is a distinct and durable informational advantage, in addition to a boon to client outcomes in the moment. This allows the platform to learn both how to support each person and how the collective firm operates. The Symbiotic Firm, as a multi-player agentic system blending humans, their personalized assistant agents, and shared firm-wide continual learning, is a whole much greater than the sum of its parts.
Because a law firm competing with the Am Law 20 is the most demanding validation of judgment, precision, and coordination we could tackle, Norm Law is the ideal refinement of the Norm platform to power high-stakes professional services firms more broadly. Through AI and human collaboration, we are unlocking nonlinear increases in firm-level productivity. But if we are rethinking the notion of The Firm from first principles, why stop there? If key high-end related professional services can can be tightly coupled for a shared client base, this would unlock more value for those clients through better coordination and more seamless service. The boundaries between service providers are as anachronistic as the workflows inside each.
This is a second-order symbiosis. At the firm level, e.g. inside Norm Law, the collective is better together with AI agents and humans fully integrated under one roof of a firm designed for this. We expect that additional adjacent firm types beyond a law firm, paired side-by-side, when using and improving our generalized underlying multi-agent platform, will in turn power the sum of the output of those firms to far exceed what each could do alone.
If you’d like to reinvent the notion of a firm in the most ambitious way possible, please apply to Norm. If our firm serves its own purpose, we can accomplish far more together than any of us could alone.