You found it! Congratulations.

What is this?

The Code of Arms Project is a new endeavor that aims to shine a light on the men and women of the U.S. military who want to see their country prevail when called to arms in the 21st century. Historically, the Department of Defense has struggled mightily with developing and field advanced technologies for the purpose of combat. The corollary, of course, is that we seek to do these things because they are necessary.

However, for many reasons, the organizational, cultural, and sociological tendency is to apply Industrial Age methodologies to Information Age problems. In a nutshell, the DoD is wired completely backwards to the Agile Manifesto.

Why serve?

Fortunately, good patriotic Americans with a strong grasp on modern technology have volunteered to serve in an all-volunteer force. They encounter headwinds upon headwinds as they attempt to apply modern technology to the problems they face. They do so while bound to rigid hierarchy, compliance with strict federal laws, regulations, and policies, and a centralized planning, programming, budgeting, and execution (PPB&E) paradigm inspired by the Soviet Union.

It is a natural and agreeable assumption that most Americans do indeed value the democratic freedoms they enjoy and the system of government that ensures these freedoms flow from generation to generation. With this lineage through time, conflicts will occur, wars will happen, and these same young Americans must stand ready to defend those freedoms.

One can argue that these freedoms should be most readily preserved on the homefront - on American territory - and thus a robust homeland defense is required. Others can reasonably argue that our freedoms are only as secure as the world around us - and thus robust power projection is required. Whether you believe in defending only the geopolitical American boundaries or in bringing liberal values to every country that will have them, those two truths remain constant: Other nations and organizations will oppose those values, and the U.S. military must be prepared to defend these freedoms, waging war on behalf of our own Constitutional values.

Fortunately for those who do not serve in uniform, young men and women volunteer to serve. The reasons are many. Code of Arms simply cares that they serve.

DoD Culture Makes Technology Hard

Zoom back in to the Servicemember. In 2023, the DoD recruited nearly 200,000 new members into its Active Duty and Reserve ranks across all Uniformed Services. At the individual level, these men and women come from the most technologically-savvy generation that has every existed. And yet, when they arrive at their duty stations, they are confronted with technologies that are partially functional, not functional at all, hard to learn, and hard to use. Many of the most sophisticated weapon systems run on software that is upgraded only once every few years - or never.

For the more enterprising types, developing their own digital solutions comes with a unique set of challenges. While the more open-ended “innovation” culture has taken root, allowing Soldiers, Sailors, Marines, Airmen, and Guardians to prototype their own answers to unit-level problems, the big programs of record - and the big primes - still dominate the DoD’s spending levels, and career field and unit training requirements dominate their daily lives. It is the rare Servicemember who is allowed or encouraged to experiment for the sake of learning. In the context of military hierarchy, there’s always a Commander’s Intent, and there’s always a mission to achieve that intent.

Suppose that Servicemember breaks through, finding time to develop code either on duty or after hours. Assuming military lawyers don’t light them up for doing Research and Development activities in a line unit not designated for traditional acquisition-type activities (like writing code), they must ensure their work meets quality and security standards in accordance with DoD policy. Of course, to conform to policy, you have to know policy, and with tens of thousands of policy documents and military technical standards, it’s simply impossible to understand the compliance space on your path to delivering a solution that is compliant with policy and ripe for routine combat or combat support.

DoD Policy Makes it Harder

Let’s talk about policy for a moment. Specifically, let’s talk about how DoD makes a relatively easy, attainable thing hard and relatively unattainable. Publicly-available generative AI can be helpful here, because it can quickly distill public policies into brief guidance. Rapidly synthesizing compliance knowledge (all those policies and technical standards) by simply asking questions to a trained Large Language Model (LLM) is an ideal use of natural language processing, understanding, and generation through generative AI. This is an area where OpenAI’s ChatGPT product excels.

But, there’s a “but,” and it’s that privately-available generative AI in DoD is a different story. The DoD is still struggling to field - let alone fund - its own internal LLM implementations. While advanced software technology (both open source and proprietary) is available to get LLM development off the ground, the hardware required for training private LLMs is in high demand and comparatively low supply. This makes it expensive. (Well, maybe not that expensive, considering today’s spending amounts and priorities.) And, you can’t share the hardware, lest your protected training data and protected models become intertwined with knowledge that shouldn’t be public.

Let’s say your unit, courtesy of the American taxpayer, lands a super-sweet high-end graphics processing unit (GPU) for training your LLM with Controlled Unclassified Information (fare thee well, FOUO). Do you have an Authority to Operate (ATO) for that system? No? Shut it down. Come back when you have one, but don’t expect any hints about who the Authorizing Official really is, and don’t expect any help with assessing each of the NIST 800-53 controls in the thousand-row Excel spreadsheet. (Yes, I know there are better ways. I’m making a point here! Throw me a bone!)

So, even if we apply modern technology to make solving modern problems with modern technology, we still end up in the Stone Ages of IT. Requirements must be decided and funding must be prioritized two years in advance (with the exception of some novel allowances from Congress, such as the APFIT program and aspects of the Adaptive Acquisition Framework) and thus many promising ideas die simply of resource starvation amid the bureaucracy. Lack of compliance - software that hasn’t been approved by some distant Change Control Board, for example - is still the fastest and surest way to kill a good digital idea.

Even if we could use modern technology like generative AI at scale, the culture itself is inherently biased against its use. This is the tension DoD finds in itself today. It knows it needs to do better as an organization, but it is hidebound by decades of Industrial Age organizing, training, and equipping, that there is scant room for modern technology in the ranks.

(Some of you are reading this, and you’re pulling your hair out because you’ve actually done what I’ve said is not possible. If that’s you, I want to interview you and tell your story. Please contact me.)

And yet…

And yet, these patriotic Americans still get up every morning, still put on their uniform, still report for duty, and still try to make the U.S. military better. There are still wars to be waged, conflicts to be fought, deterrence to be maintained, and power to be projected. There are still freedoms to be prepared to die for.

The Code of Arms Project aims to tell the stories of these Americans in uniform. It aims to capture their challenges and how they’ve overcome them. Code of Arms is about the Servicemembers and how they endure. It’s about how they persevere, even when “the system” is seemingly working against them at every turn, year after year. It’s about the champions who find problems they can solve get their friends to try the solution they’ve found.

Code of Arms about the infantry officer who’s willing to walk across the FOB to build a relationship with the spectrum manager so their radios perform correctly. It’s about the nuclear submarine engineer who sees enough merit in upgrading a software package that they forge the path to upgrade it. There are many stories like these. The U.S. military needs to change the way it fights, and the people highlighted here at Code of Arms are the torchbearers, the trailblazers.

Code of Arms is not about PPB&E or acquisition reform, ATO process improvement, or hybrid Communist-capitalist surveillance states. There are really bright people who have much more useful things to say on these topics.

Code of Arms fills a gap in the national narrative. It speaks to the Herculean and Daedalean efforts of technologist pipe-hitters who make DoD better from the inside out, even if doing so isn’t the easy way.

We’re here to tell you their stories. We hope you enjoy them, and we hope you’re inspired to support them, back them, and encourage the change they know is so, so necessary. Their experience, knowledge, and perseverence is what this nation needs to fight and win in the years and decades ahead.

These views are my own and do not necessarily reflect those of the Department of Defense.

Author’s Note: This project originated from a conversation with my wife two years ago. I had just finished reading James Patterson and Matt Eversman’s “Walk in My Combat Boots” and several years before had read the eponymous IT novel, “The Phoenix Project.” I felt there was room for a project that combined the best of both: A blueprint for organizational IT transformation, and telling the gritty, so-not-Hollywood stories from inside the U.S. military. If anyone gets the credit for this project, it’s actually her. I’m simply the messenger.