Patterns of Conflict: Behind the Scenes
My story of tragedy turned into purpose, and the discovery of the OODA loop
I'd like to ask you a question: what's the purpose of a hacker?
By the end of this writing we will a different frame of reference in which to discover that answer.
The Nature of a Hacker
I am driven to research the character and nature of being an offensive operator, as part of my journey into answering the question above.
I've read about and studied people like William Sherman, Scipio Africanus, Marcus Aerelius, and Naval Admirals; anything that I could get my hands on that I could try to adapt to offensive cyber operations, I put on my list.
During my research, I discovered a man by the name of Colonel John Boyd. Specifically, Col. Boyd’s famous "to be or to do" speech, intrigued me. I picked up an audiobook on John Boyd and started listening and what I discovered was that he too went on a journey of discovery not on the character and nature of war but human conflict in general.
I resonated with Col. Boyd's journey and philosophy, as just this past year I presented a talk at a couple of hacker conferences on how naval fleet tactics are applicable to cyber operations. I’m out to get people to look beyond the thin veneer of cyber operations to find reskinned version of the character and nature of war that humans have been redefining since the start of civilization.
Alive on Paper Only
John Boyd lived before my time. He gave presentations and briefs in classified environments like the Pentagon, places where secrets don’t leak (←irony). One of his most famous briefing was titled Patterns of Conflict, research he had redefined over many years.
I had zero expectation that what I had read about would be visible to someone like me on the outside, especially something like Patterns of Conflict (called simply ‘Patterns’).
Imagine my surprise one day when googling for info on Patterns, I discovered that another Boyd acolyte (and I do consider myself one now, based on the sheer amount of time I have spent listening to this man's voice), discovered in the congressional archives a recording of John Boyd briefing Patterns of Conflict to some staffers.
I was blown away, because the people I read about exist only in books: in history. But here was John Boyd speaking, moving, and giving the presentation that I was certain I would never get to experience.
Provide Solutions, Not Problems
As I watched the video I discovered something: it was essentially unusable, unconsumable in it's current format.
Why?
The video was grainy, choppy, and only the audio was distorted and only coming out of one side of my headphones. While I technically could power my way through the presentation, I would spend so much mental effort translating everything I was watching and hearing into something usable that I knew I would need to watch it multiple times.
And I don’t think I’m alone, as the drop off after the first video is staggering, with the appearance of most people not making it to the final video.
So here’s the problem statement: I found the video of something I never expected to find, but it’s unconsumable. As Col Boyd says often, “Let's put a pin in that for a minute and we'll come back to it.”
Part two of this backstory involves my dad, and this is the first time I've brought up the story I’m about to tell publicly.
From Tragedy, Purpose
My dad passed away in April of 2022. I wasn't close with my dad, but he was still my dad. During the week leading up to the funeral, I spent a lot of time with my dad's side of the family, bonding with them more in that one week than in 30+ years combined.
My uncle found this old VHS tape of he and my dad playing together at a rock concert. But it, like Patterns of Conflict, was in an unconsumable format.
Being that I am part of company that specializes in AI, I figured there might be some way I could contribute? Could I "clean" this VHS tape of my dad and make it better? The answer was yes. Yes I could, but I head to learn how to first. And quickly.
Over the course of the next couple days I poured myself into some of the tools my company has released, Maxine, to restore, clean or really just tinker with my dad's VHS videos.
I also had to make a finished product out of it too, not just upscale and go. So I had to learn a video editing program as well, like DaVinci Resolve.
So that's what I did: I got to work and learned. I spent around 20 hours over the course of two days figuring out how to upscale the audio and video of the VHS tapes and put them into DaVinci Resolve to make this final tribute to my dad for his celebration of life that week.
And regardless of how the video turned out, I got to spend a couple more days with my dad. That's how I see it at least. I spent hours watching his movements, listening to his voice, zooming in on his face to highlight his movements.
And those are precious memories. Precious and unexpected.
Patterns of Conflict
So can we start drawing conclusions about how the experience with my dad's videos led to Patterns of Conflict? I think so.
Part 1: Here was a man, John Boyd, who spent years trying to get people to listen. Who had a drawer full of uncashed checks because he didn't want to compromise his message by "selling out." And while some people did listen to his message, I don’t think enough listened to make him feel like he had fulfilled his life's purpose (my conjecture).
Part 2: I had picked up skills just months before, and I had just discovered a video of someone who I never thought I would hear or see, talking about something I also was deeply passionate about, with a problem similar to my dad’s videos.
Part 3: when I was in the army I was a linguist. I spent hours learning how to transcribe, a skill that I thought would never come in useful.
By combining all three parts, I had a plan and a way to help preserve John Boyd’s Patterns of Conflict briefing: cleaning up the audio, cleaning up the video, transcribing the presentation in total so that those who can’t listen to it can read it.
So if that’s the how? Then what’s the why?
Why do this?
I'm doing this for a few reasons.
First, I believe John Boyd is the closest thing we have to a modern day Sun Tzu. To be able to hear and see Sun Tzu, that’s an lifetime opportunity to me. And much like many others preserved Sun Tzu, Col Boyd’s work needs preserved too. And I can help do that.
Second, I plan on adapting Patterns of Conflict to Cyberwarfare, something I was already researching and presenting. But to do that I first need a basis of which to work. Which means I need to make Patterns of Conflict consumable for at least myself. And if I can make it consumable for myself, I can make it consumable for everyone else.
And third, well if there is an afterlife, maybe John Boyd will see that others found his message important enough to preserve and spread. Maybe my work will help provide a sense of closure to that one aspect of his life.
And maybe, help provide some closure to an aspect of mine.
Emotional stuff aside, here's the process for preserving John Boyd.
The Technical Stuff
Get the videos, there are 13 (Pic above)
Download the Maxine audio and visual SDK (link above)
Video
Denoise and remove artifacts from the video (the audio is stripped out with this step)
DenoiseEffectApp.exe --model_dir=..\..\bin\models --strength=1 --codec=H264 --show --in_file="Patterns of Conflict Part 1.mp4" --out_file=PoC1_dn1.mov --progress
VideoEffectsApp.exe --model_dir=..\..\bin\models --effect=ArtifactReduction --mode=0 --codec=H264 --show --progress --in_file=..\DenoiseEffectApp\PoC1_dn1.mov --out_file=PoC1_DN1_AR0.mov
Upscale the video using the video SDK (much trial and error - you can only polish a turd so much before it’s a high resolution blurry turd)
VideoEffectsApp.exe --model_dir=..\..\bin\models --effect=Upscale --strength=1 --resolution=720 --codec=H264 --show --progress --in_file=PoC1_dn1_ar0.mov --out_file=PoC1_dn1_ar0_720UP1.mov
Audio
Fix PATH
SET PATH = %PATH%;C:\Program Files\NVIDIA Corporation\NVIDIA Audio Effects SDK
Strip out the audio from the source video using VLC.
Convert the MP3 to a WAV using a website. To use the audio SDK full features, change the channel to mono, the bit rate to 16, and the sample rate to 16k.
Very light denoise and dereverb, then upscale the audio to 48k
first: denoiser16k_cfg_ampere
effects_demo.exe -c denoiser16k_cfg_ampere.txt
second: dereverb16k_cfg_ampere
effects_demo.exe -c dereverb16k_cfg_ampere.txt
third: superres16kto48k_cfg_ampere
effects_demo.exe -c superres16kto48k_cfg_ampere.txt
You have to edit the txt files directly that the effects_demo.exe uses
Example directly below for denoiser16k_cfg_ampere
# Effect
effect denoiser
# Point this to the model file
model ..\..\bin\models\ampere\denoiser_16k.trtpkg
# Noisy input file
input_wav boyd1_16k.wav
# Denoised audio data will be saved to this file.
output_wav boyd1_16k_dn_08.wav
# Set to 1 for real time mode i.e. audio data will be processed
# at same speed like that of an audio input device like
# microphone. Since the denoising is faster that real time, the
# processing will be equal to audio file duration.
real_time 0
# Intensity Ratio
intensity_ratio 0.8
# Enable VAD
enable_vad 0
Transcript
I use descript for my videos anyway, so I dropped the audio file into descript and let it do an initial pass. It’s better to start with something rather than nothing: it’s that blank canvas analogy I always go on about (ask the people I work with how many times they have heard it).
Start listening and changing words. 10 mins of audio time takes between 90-120 minutes of real world time. This is due to the low quality of the audio and that Col Boyd isn't using a microphone and often speaks away from the camera.
For tricky or confusing words, I had three options: search through the project white horse presentation, the transcript of patterns that was given to the Marines in 1989 (Col Boyd often changed his presentation, so there was no one source of truth for tricky or hard to hear words), or sometimes I just had to straight up google something that sounded close to what I thought it was and hope. I found most of them, but not all.
Once the transcription for the video was complete, I take the audio and video and put them both in DaVinci Resolve.
Because of the video upscaling, the video is now longer than the audio. To line them up I had to do some napkin math and adjust the video speed so that it's as close as possible to the original length.
I add in the subtitles from Descript and line them up. Then I watch the entire video again and pick out mistakes I made with the transcription, go back and fix it, and then put the fixed transcript back in. Nonverbal communication can lead a person to the correct words they were missing, so remember that the next time you are in a call with your video off!
Next I take the video w/ the subtitles and start marking where slide transitions are and add the slides at those points. By now I have listened to or watched this section of video at least 10 times, but finally I can really start to study it and understand it, not just listen to the words.
As I go through the almost completed video, with the slides, and look for key points I note them down. I look for natural break points in the instruction, and as an example the first video released in this series is all of the first original video and a few minutes of the second video added at the end.
Part 1 - Moral Mental Physical
In Part 1 of Patterns of Conflict, Col Boyd talks heavily on the categories of conflict, and how they play into conflict with an adversary. Where I stop the video is where he transitions to another lesson to expand upon those ideas. Remember, I’m trying to increase the consumption of Col Boyd’s work. Laying out a 4 hour video isn't going to help with that.
To see the final product of ALL those steps above, check out Part 1 of Patterns of Conflict: Moral, Mental, Physical.
What’s next?
So what's next? At the time of this writing I have 7 parts of 13 transcribed with the slides on the video; I'll be releasing those probably on a weekly basis.
And to further help with consumption of the lessons, I’m going to make audio transcripts for each slide to function almost like a podcast, using Descript to release AI generated voice for each slide.
Some people want to watch, some people want to listen while they drive, walk, exercise, whatever. Unfortunately, you can't listen to Col Boyd's video while you are driving because the transcripts are needed (unless you've watched the specific video about 10 times). So we'll do that, take the transcript and make a podcast out of it for each slide.
Beyond that, I'll get back to transcribing the rest of the vids, getting them out there for the world.
And even beyond that, then I can finally start working on my application of patterns to the nature of cyberwarfare, dubbed either the Operator Weapons School, or the Cyberfighter Weapons School (I’m still deciding on the title). But the goal is to be a war college for cyber operators, to progress through the lectures at various stages in their career.
I’m not going to teach tools and how to exploit. I’m going to ask people to think differently about what it means to be a hacker: whether just starting out, a hacker that leads a team, or a hacker that leads a company. We will explore how best to accomplish the goals of a hacker using our unique talents and methods. I'm very excited to get started on that.
Stay tuned and thanks for reading!