The Department of Defense’s vision of Combined Joint All Domain Command and Control (CJADC2) is right. Realizing that vision requires we first ask ourselves two questions: Are we organized to meet that moment? If not, how should we organize around it?
The biggest obstacle to achieving JADC2 is not technical, it is organizational. In 1967 computer programmer Melvin Conway coined Conway's Law which stated that "any organization that designs a system will produce a design whose structure is a copy of the organization's communication structure." This is the discriminator on how software is built and the substantive reason Silicon Valley companies fetishize organizational structure (their product and market competitiveness is a literal consequence of it).
Conway’s Law can be understood through this lighthearted and extreme caricature of some of America’s largest tech companies. Even as a caricature, it roughly explains why different companies succeeded or failed in the face of different disruptions and opportunities. Microsoft in the early aughts was too busy fighting internally to seize the opportunity presented by mobile (ceded to Apple and Google) or cloud (ceded initially to AWS). Google is so matrixed that it moves slow and couldn’t seize on Cloud even though they invented Kubernetes, map reduce, and basically all of the foundational technologies of the modern cloud. Apple is functionally a series of Special Access Programs tightly managed to create exquisite, discrete offerings. Meta is a flat web of peer to peer experimentation and innovation. And so on.
Conway noted that when various organizational divisions or groups within a bigger organization collaborate on a more extensive system, they would inevitably divide the system into components that each team could work on autonomously. Subsequently, they would determine how these individual systems would interact with each other via a specific communication method, resulting in an end product that resembled the organization.
Funding services to drive progress via Army’s Convergence, Navy’s Overmatch, and Air Force’s ABMS is likely to result in value in many conceivable ways, but perhaps not in JADC2. This is a problem since at a fundamental level the Department is organized to execute programs through the services. The risk of Conway's Law manifesting in JADC2 is that each service's efforts will prioritize their own unique requirement and communication structures, leading to a disjointed or... non-joint solution. This would undermine the very essence of JADC2, which aims to provide a unified and integrated command and control system across all domains and services.
It seems like DepSecDef Hicks realizes this possibility and has therefore taken on an outsized role with the AI and Data Accelerator Initiative (ADA) and the Chief Digital AI Office (CDAO). We’re also seeing CCMD commanders realize that JADC2 comes together at the CCMDs or not at all - hence the big investments being made by Gen Kurilla with CENTCOM’s Digital Falcon Oasis and by Gen VanHerck with NORTHCOM’s Global Information Dominance Exercise (GIDE). That JADC2 likely requires a CCMD-centric approach is further reinforced by DIU’s focus on delivering capabilities to the CCMDs via its Hedge portfolio. Chris Brose’s Moneyball Military paper argues that, “…the services cannot monopolize the government’s role in the supply of military power. More of that function must be shifted to alternative institutions that report to the Pentagon’s top leaders, to whom they should be accountable for disrupting the status quo and supplying Moneyball capabilities directly to military consumers.”
If the services are going to drive this, to avoid the pitfalls of Conway's Law and ensure a truly joint JADC2, the focus should shift from individual service efforts to a Combatant Command (CCMD) centric approach. This would involve each service's JADC2 efforts competing against one another to win the support and endorsement of each COCOM. By doing so, the services would be incentivized to prioritize joint requirements and interoperability, ensuring that the final JADC2 solution is truly integrated and capable of meeting the needs of all stakeholders.
This COCOM-centric approach would not require a complete reorganization of the DoD, as a Goldwater-Nichols 2.0 might. We have two options:
Leverage the existing organizational structure and promote healthy competition between the services (with funding tied to what COCOM commanders chose to fight with) to achieve the desired joint outcome. Each service has to win all the Components and the CCMD itself at a given COCOM.
Decide that the software for a joint force can’t be built through the services due to Conway’s law and charge DIU or CDAO to build and deliver this. The ideal outcome would have at least two competing efforts. Over time the services will retain equipping their forces in all but software. In software they will compete against each other to break the DoD monopsony and drive continuous competition and speed of war innovation.
Reorgs cause a lot of brain damage, confusion, and wasted time, when we have none to waste. JADC2 is fully realizable if instead of reorganizing our Department, we reorganize around the problem.