I believe the process is repeatable and thus far has produced useful outputs, at least for my work. I’d like to share it with you, not only to hopefully verify its usefulness across a more broad spectrum of use cases but also in the “open-source” spirit that made me fall in love with software development in the early 00s.

I’m calling the process “GAN-coding”, in contrast with the widely used and overloaded term “vibe-coding”.

What is GAN-coding? (and what it isn’t)

Vibe-coding is a slangy term used to describe a style of software development in which a human uses natural language prompts to instruct an AI model to generate code rather than manually writing code by hand.

GAN-coding is a software development methodology where code is produced through repeated adversarial cycles between generators and discriminators, with humans retaining final authority, and correctness is enforced through explicit rejection loops rather than trust.

This isn’t “GAN” in the machine-learning sense, technically. I’m borrowing the generator/discriminator pattern because it matches the workflow: produce an artifact, then adversarially challenge it.

“Vibe-coding” has a broad scope. In practice it can mean hyper-rigorous AI-assisted coding very much like what I’m labeling “GAN-coding”, OR it can mean a one-shot “make me a website that takes payments for a digital product” or anything in between. When I refer to vibe-coding in this article, I’m scoping down to “average-ish” entrepreneurial or professional use of AI-assisted coding.

GAN-coding can be seen as an extension of vibe-coding with some key differences:

It’s like the “TDD” (test-driven development) cousin of vibe-coding. If you already do strict code review + tests with AI assistance, GAN-coding is the largely the same but I’ve attempted to make it explicit, role-driven and repeatable.

Core Principles

1. No Single Agent Is Trusted in Isolation
   Every meaningful artifact—design, code, tests, or review—is challenged by an independent agent. Humans and AI models alternate roles as generators and discriminators.

2. Discrimination Is the Bottleneck
   Generation is fast and inexpensive. Evaluation, understanding, and rejection are not. The process is intentionally optimized to preserve human attention for high-leverage decisions.

3. Correctness Is Proven, Not Assumed
   Code must survive adversarial scrutiny: tests that actually verify behavior, reviews that look for failure modes, and diffs that can be reasoned about line by line.

4. Diversity Defends Against Correlated Failure
   Multiple AI models are used intentionally, not interchangeably. Different priors, biases, and blind spots reduce the risk of silent, shared error.

When we adopt these principles and the process I’ll describe below, some interesting things happen.

If the human doesn’t fully understand the code, then the human cannot be accountable, or has an exit hatch for accountability. If rigorous validation is optional, as is the case in “average-ish” vibe-coding, then failure emerges from false confidence. “The AI scores high on SWE so this must be good to go.”

While far from perfect, human manual coding was the progenitor of most of the Internet up until recently. It’s the devil we know and we have process around it. I propose we apply similarly rigorous process to AI-assisted coding in order to reap the considerable benefits AI can confer while maintaining some semblance of human ownership, accountability and authority that helps align outcomes with intent and value.

The GAN-coding Process

A few critical features of the GAN-coding process:

• Explicit human rejection loops and human accountability

• AI-assisted adversarial discrimination cycles driven by human judgement

• Constraint-driven prompting (tests, invariants, contracts)

• Human-owned architectural decisions

Phase 1: Design

The process begins with traditional requirements gathering, specification writing, and architectural design. This phase is intentionally aligned with best practices from manual human coding. The human generates design artifacts, specifications, requirements, etc.

Roles

• Human: Generator

• AI (design-review model): Discriminator

The AI is used to challenge assumptions, surface edge cases, and critique architecture, not to author it. Design concerns are resolved before implementation planning begins.

Phase 2: Implementation Prompt

A canonical prompt is produced that describes the system with sufficient clarity that either a human engineer or an AI system could implement it.

This constraint is deliberate. If only a specific model can interpret the prompt correctly, intent has leaked into the model rather than being captured in the design. In GAN-coding, prompts function as contracts, not vibes.

Roles

• AI (design-review model): Generator

• Human: Discriminator

Phase 3: Implementation Planning

A coding-focused AI model decomposes the work into explicit phases or chunks, ideally with semantic and functional boundaries. The human reviews and approves the plan. Coding does not begin without an approved plan.

Roles

• AI (coding model): Generator

• Human: Discriminator

Pro tip: keep iteration phases small to prevent context drift.

Phase 4: Iterative Code Generation and Testing

For each phase:

1. AI generates code.

2. AI generates a test suite targeting that code.

   1. Could do #1 and #2 in reverse order for closer TDD alignment.

3. The human scrutinizes the tests:

   1. Do they verify behavior or merely exercise code?

   2. Do they cover critical paths and error modes?

4. The human reviews the code for comprehension, targeting medium-high understanding. Adversarial tests and independent reviews act as safety nets for the parts the human hasn’t fully parsed, but tolerate gaps with extreme caution.

5. All changes are committed. At every step, diffs are reviewed manually by the human.

Roles

• AI (coding model): Generator

• Human: Discriminator

Iterations continue until the phase converges.

Phase 5: Independent AI Review

A different AI model performs a full review of the code and tests. The human is responsible for selecting a discriminator model suitable to the task and providing sufficient context for an effective review: requirements, prioritization, convention, domain knowledge, etc. NOTE: this happens before pushing a PR for automated AI code review in CI/CD.

Roles

• Original AI: Generator (as source of artifacts under review)

• Independent AI: Discriminator

This phase defends against human fatigue and blind spots but also injects diversity of thought to reduce correlated failures.

Phase 6: Review Arbitration

The human reviews the AI reviewer’s feedback. Valid concerns are verified by the original coding model.

Roles

• Independent (reviewer) AI: Generator

• Human + Original (coding) AI: Discriminator

While the introduction of the original coding AI as a discriminator may introduce defensive rejection (”my code is solid”) the original coding AI possesses unique context that can greatly reduce false positive issues. There’s an implicit additional step in which the roles are:

• Original (coding) AI: Generator

• Human: Discriminator

Only verified issues are addressed. All subsequent changes are subject to “Phase 4: Iterative Code Generation and Testing”.

Completion and Convergence

The system is considered complete only when:

• All planned phases are implemented

• Tests meaningfully enforce invariants

• The human can explain the system at an architectural and code level

• No unresolved discriminator objections remain

Recursion: Verified changes loop back to Phase 4. Upon phase completion, the cycle repeats for subsequent phases until the project is complete.

Subscribe now

In Practice

Is this actually useful? Isn’t it boring and expensive?

Yes. Yes.

For many software engineers, individuals or teams, this process may net out to reduced productivity. At first. But there’s a similar argument to be made for TDD in general, or for the stricter quality controls that are natural extensions for scaling, successful products. The argument is that the cost of rigor gets amortized, paid off, and starts returning value over time. (And by the way, some or all of your team members may be doing this already.)

That said, the GAN-coding process as it stands isn’t cheap. To calibrate on when this is worth the cost, let’s posit some dimensions on which we can gauge when GAN-coding could net more valuable returns.

For small teams and prototype products, GAN-coding might be a time-sink with minimal return. Say “no” to GAN-coding during rapid prototyping, or in the cases where tests are disproportionately expensive relative to the risks they mitigate.

If speed is practically irrelevant, the codebase super-stable and mature, or there’s little pressure for engineers to stretch beyond their capacity, then manual human coding may be the best approach that introduces the least risk. But even then, GAN-coding can provide valuable diversity of thought to detect otherwise hidden failures and tacit assumptions. When the iterative review cycle and AI assistance inherent in GAN-coding results in more thorough documentation or better test coverage, any team could benefit from improved maintainability.

The Human Element

GAN-coding isn’t for everyone. It doesn’t cater to the strengths and weaknesses of every coder equally. Let’s look at a few skills that may be relevant (this list is far from exhaustive).

“Average-ish” vibe-coding rewards expressive intent over technical depth. You don’t need to comprehend the outputs. You “load it in your browser” and if it works, it works.

Manual human coding rewards deep expertise, it sometimes rewards an architectural mindset and skepticism, it demands (but doesn’t always get) a high level of debugging and failure analysis, and there’s a manual dexterity component that makes it uniquely human (for now).

GAN-coding rewards judgement and curiosity.

No Single Agent Is Trusted in Isolation.

This core principle requires the GAN-coder to not only be skeptical of the outputs of others, but also their own outputs.

Discrimination Is the Bottleneck

The GAN-coder, by virtue of selecting the methodology, is under strain to extend their innate capacity. Their curiosity drives the resilience required to review yet another line of AI-generated code and also to learn from it. It’s only boring if curiosity is exhausted.

Correctness Is Proven, Not Assumed

The GAN-coder is required to set aside their biases, practice objective judgement and be curious about the truth rather than their preferences.

Takeaways

While GAN-coding may not work for everyone, I posit that to those it does work for, it will be greatly beneficial. Here’s a recap:

• No agent trusted in isolation.

• Write a spec a human could implement.

• Don’t code without an implementation plan.

• Correctness is proven. Use tests and always review them manually.

• Discrimination is the bottleneck. Always run an independent model review before PR.

• Diversity defends against correlated failure. Use different models for different roles.

• The human is always the tiebreaker.

Note there are some nuances in terms of prompt engineering, model selection and other bits I collectively consider “implementation details.” A goal of mine for the GAN-coding process is that it can be successful agnostic to such details, as long as each step of the process is implemented faithfully. The one exception is the prompt in “Phase 2: Implementation Prompt”. That’s a detail I feel worth reiterating.

If you’ve read this far, I think it means something resonated with you. I’d look forward to your feedback and input. This is a work in progress, after all. I used something like the GAN-coding process to write this article. I produced outputs that I asked three different frontier AI models to review, validate, and enhance. I didn’t write the mermaid diagram by hand, but I edited and iterated on it with an AI collaborator. I pasted the entire output for final review by multiple LLMs.

The final authority and discriminator though, is you :)

Does everything seem more difficult, lately?

I'm with you. Work, world news, personal life, finances, anxieties about all those things, second-order anxiety about those anxieties...it's a lot to carry.

Recently I came across a concept that, in the context of DnD1, is intuitive and easy to wrap my head around. This makes it a surprisingly useful tool for managing stress.

The concept is based on resilience. Michael Rutter, often referred to as the father of child psychology, describes resilience as:

It makes sense that if we lose our ability to bounce back from difficulties, we become more susceptible to emotional and mental disorders. We can think about "disorders" as a "spectrum of disorder". We're not necessarily talking about a diagnosis like Generalized Anxiety Disorder, but rather a more generic "lack of order".

Important: I'm not equating stress overload with diagnosed disorders. This article is a framework for thinking about stress and resilience, not a substitute for professional help. If you, or someone you know is struggling with mental illness, please seek guidance from a licensed professional.

Resilience is a way for us to stay on the less chaotic side of the disorder spectrum. The ability to bounce back from life's difficulties is universally desirable. It follows that we would want to improve or strengthen our resilience. But how?

If there's any truth to the adage, "you can't improve what you can't measure" then we need a way to quantify our resilience. Then we need ways to improve on it, and ways to measure that improvement. We need to track our progress (and our regressions) over time, in order to make adjustments and introduce more improvements.

This is where DnD comes in :)

DnD gives us a few ways to model resilience conceptually. A character's AC or "armor class" represents resilience against physical attacks, and is based on the character's level of "Dexterity".

But thankfully, physical attacks aren't the most common stressor in modern life, unless you're in a few specific professions that we'll give our respects to here but not really focus on.

We're talking about emotional and mental resilience. There are other ability scores in DnD that might fit the bill: "Charisma" or "Wisdom" for example. "In-Dn-Deed" those abilities endow a character with resilience against some non-physical threats, like being persuaded to do something you don't really want to do.

We're looking for a more broadly applicable type of resilience. And perhaps a more interesting connection. And one exists, if we examine the ability score, "Strength"—although not for the reasons you might think.

If you've played Baldur's Gate 3, you've likely noticed that at a certain point in the game, you start spending an inordinate amount of time looking through your inventory—the list of things in your knapsack. You start right-clicking and choosing to "Send to Camp" anything that isn't immediately relevant to what you're doing in-game. You don't just carry it all around with you all the time.

That's because every character has a Carrying Capacity, which as it happens, is based on their level of Strength. More Strength, more Carrying Capacity. When a character in DnD picks up an object, that weight is subtracted from their Carrying Capacity. When they reach the limit of their Carrying Capacity, they start to do everything worse. Even walking becomes slower. Every action becomes way more of a chore when your character is over-capacity. Eventually they can carry nothing more. The DnD term for this condition is "encumbered".

In the regular course of our lives, we encounter stressors of various intensities. What if we gave each of them a "weight?" A hard deadline at work might be a "20". A hard deadline that you feel is unreasonable or unachievable might be a "40". Moving house is considered one of the most stressful experiences in life, and maybe that's a "90", above which is only death in the family and critical illness, "95" and "100" respectively.

In that sense, we're starting to think about our ability to bounce back from life's challenges and unforeseen difficulties as our Carrying Capacity2. There is a number that represents our resilience, and a number that represents every kind of stressor you could encounter. Some stressors "stay in your pack" while others you can unload every night at camp. When you reach the limit of your resilience, you start to lose protection from emotional and mental disorder. You become encumbered.

Just like Carrying Capacity in DnD, our resilience number is more or less fixed, on a day-to-day basis. Also like DnD, there are ways to increase our Carrying Capacity, but these happen over longer periods of time. Right now, you can't pick up that barrel of "runepowder" if you're also carrying a bunch of plate armor. Long-term, you can level-up and increase your Strength, but short-term you need to unload some stuff.

Thinking of resilience in this way goes beyond the conceptual to giving it a numeric value and a hard limit. Thinking about life's challenges in the same way means we can assign numbers to those, as well. Armed with simple math, it's straightforward to assess your current load.

Let's say your Carrying Capacity is "100". If you're moving house right now, carrying a "90" in "weight", you're probably going to struggle adding on a hard deadline at work.

I didn't really believe it when I first came face-to-face with the reality that there exists a limit to my ability to deal with life's challenges. We all deal with challenges, and we all just keep going, right? As humans we are more adaptable than our in-game avatars.

But when I think about a DnD character's encumbrance, it starts to make sense. Sure, they keep going, but things get bogged down. Similarly, as you and I become encumbered, everything we do will be negatively impacted.

• It might take longer to complete each task.

• You might make poorer decisions in terms of prioritizing tasks, orchestrating complex processes, or minimizing opportunity costs.

• You might get distracted and reach for cheap-dopamine distractions, because your emotional control is weakened.

• You might get irritable, affecting not only your work relationships and therefore your ability to get things done, but also your personal life.

• You might neglect important self-care routines, especially if you work from home.

• Your quality of sleep might suffer, adding yet another stressor.

• You might make poor nutritional decisions due to lack of time or emotional control, which in turn can impact your resilience.

These things happen whether you realize it or not. They can be self-propagating, eroding your resilience in a vicious cycle. This can lead to catastrophe, making you more susceptible to making mistakes at work, for example. When your mental faculties are stretched to their limits and beyond, it's no wonder you might experience "burn out", or worse. With this framework, you can make assessments, even daily.

Even so, sometimes life throws a lot at us, and no amount of math is going to make it all go away. We need ways to manage encumbrance when it occurs. We need to measure AND improve.

We already talked about how, in Baldur's Gate 3, we'll often unload excess baggage to the "camp chest". Identifying stressors in our lives that can be deferred, or unloaded at least temporarily, is a solid practice. Having a maximum number to work with helps us prioritize what to unload.

Take for example, Warren Buffet's "25/5" rule:

While ruthless, it's critical to have this kind of prioritization strategy or something similar in your toolkit. Prioritization is a broadly covered topic in the realm of productivity and full coverage is out of scope for this article, however searching online can help you uncover many strategies, frameworks and skills to learn.

After prioritizing and pruning your inventory, there's another lever we can pull. Even though increasing our resilience is a longer-term play, we have a short-term option, which is to solicit more resources.

In terms of emotional and mental resilience, a resource is any tool, technique, practice, or support system that increases your ability to cope. By no means exhaustive, here's a list of examples in no particular order:

• The 4-7-8 breathing technique: a quick and simple way to signal your body to relax, by slowing down your breathing and activating your parasympathetic nervous system.

• Mindfulness and meditation: a practice of being more present in the moment, which can help reduce anxiety about the future or regrets about the past, while cultivating a more peaceful mind.

• Practices that promote self-awareness and gratitude, like journaling: writing down your thoughts or listing what you're thankful for helps shift your perspective from stress to appreciation, and fosters personal insight.

• Maintaining an exercise routine, or at least spending active time outside: physical activity boosts endorphins, improves mood, and directly increases your overall resilience to stress.

• Cultivating friendships and family relationships: staying connected with supportive people helps counteract isolation and provides a sense of belonging, both of which are crucial for emotional well-being.

• Doing something creative or inherently rewarding: hobbies like painting, cooking, or playing music help refocus your mind, boost self-esteem, and offer a refreshing break from everyday pressures. I find wheel-throwing pottery especially grounding.

• Seeking professional help when needed: talking to a counsellor, therapist, or other mental health professional can provide personalized strategies and support—sometimes, you simply need expert guidance to navigate life's toughest challenges.

To further our DnD metaphor—perhaps beyond its welcome ;) —a helpful resource might be a party companion to whom you can pass on your extra baggage, casting a spell that enables you to carry more, or buying a magical "bag of holding".

In a 2016 paper published in the Journal of Sport Psychology in Action, David Fletcher and Mustafa Sarkar, "presents an evidence-based approach to developing psychological resilience for sustained success." This implies an active process of leveraging and cultivating our internal resources, which can be developed over time.

Just as you can level up a DnD character's Strength, you can cultivate your own resilience with habits that strengthen your mind and body over time. Practices like these can have short term impacts and also contribute to your baseline resilience for the long term: a win-win.

While on one hand we can work to increase our innate resilience, I've personally also found a lot of benefit in shifting my perspective: thinking of resilience as a limited and precious resource.

I used to just take on anything and everything with no regard for what it would mean to the rest of my life. I still find myself doing that. But I've started to become more aware of each additional stressor, the weight of each item I put in my inventory. I've started to be okay with having limits, and taking care of my resilience as a valuable and critical component of my well-being.

Life can laden you with heavy loads, but you don’t have to carry it all at once. Delegate what you can, level up your resilience, and remember—you’re not alone on this quest3.

===

Here's an example table of "weights". It's only an example. Everyone experiences stress differently and the "weight" of any given stressor will be different for different people or at different times. Yours will be specific to you, based on your values, situation, and inherent strengths or weaknesses in dealing with certain kinds of stressors.

For simplicity I use a maximum Carrying Capacity of 100. Your number may be different. Just remember to assess the values of your resilience and stressors under the same lens, at the same scale. When the sum of your current stressors approaches your limit, it's time to offload, prioritize and seek additional resources or support.

Some folks think depression is about pain. Lack of motivation. A spell or phase to “snap out of”.

Sometimes they're full of love, a deep desire to help. Sometimes they're full of contempt, full of righteousness. Sometimes they see a depressed person as a problem to be fixed.

These folks don't understand. They can't. If they did, they couldn't love, couldn't help. They wouldn't know contempt nor righteousness, for both require engagement. To see a problem worth fixing is to see anything worth anything.

No, depression isn't about pain. Pain would be a relief.

Depression isn't a problem. A problem would be interesting.

It’s nothing to snap out of, because for the depressed person there’s nothing to snap into.

Imagine you are the lone inhabitant of a small asteroid. It has all the life support required for you to stay alive indefinitely, floating alone through space. The sky is featureless, a blank canvas of nothingness. No stars appear, because a nearby Black Hole consumes all light. It will consume you as well. You and your asteroid drift slowly towards the inevitable destiny, that is to be crushed into oblivion. The anticipation means you get to experience being crushed into oblivion right now. And the end will simply be more of the same.

Depression is The Void. The never-ending abyss.

I remember what it felt like, drifting in The Void. If you can feel anything at all, you feel trapped. There's no escape from a Black Hole.

People might send messages to my asteroid. They might tell me:

“Things will get better.”
The Void offers no evidence whatsoever that this is even possible.

“Take care of yourself.”
Why?

“Seek help.”
Sound advice.

But if help isn’t even conceivable to me, what would have to be true for me to seek it?

Someone would have to visit me on my asteroid. They would look around, and see for themselves. They would feel the hopelessness too—but they wouldn’t be hopeless.

Instead they would keep visiting. They might not say much—there’s really not much to say. Maybe they’d hold out their hand, and I wouldn’t take it, because why? But if they kept doing it, without pressuring me, just dropping by, and seeing for themselves, and holding out their hand. If they kept doing that, then maybe one of those times, I might take it, because why not?

If you’re someone who loves someone in The Void, you might feel pretty helpless yourself. Know that you matter. Simply being there can be the only thing that does matter.

I remember what it felt like. And if you’re somehow reading this from that place, I'm here to say to you:

The Void is real, but it is not inevitable.

There is a way out.
It’s possible to find it.
I did.

I hope you find it too.

---

This article is a personal reflection, not professional advice. Depression is a serious and potentially life-threatening condition. If you or someone you know is experiencing depression, please reach out to a mental health professional or a crisis hotline for support. Help is available, and you are not alone.

We chose a conference room a little ways from the main office area. Tom fiddled with his hands as he found a seat. I closed the door and sat across the table from him. Tom was one of our best new hires, and if he was having a hard time, I very much wanted to help.

Subscribe now

Tom described how he’d been working for weeks on his assigned task. He finally submitted his code for peer review, only to be rejected. The reviewer asked for a lot of changes.

“It’s nothing to worry about, Tom. Making changes after code review is just part of what we do. If you need help we can schedule a pair-programming session.” I was poised to solve the problem right there and then with my knowing advice and steady guidance.

Luckily, before I spoke, Tom’s facial expression reminded me of an overlooked superpower for supporting others.

—

“What if I told you, ‘right now’?”

I could hear the soft rustle of clothing as my fellow audience members adjusted themselves in their seats.

I glanced around, and saw heads slowly nodding. A woman sat contemplating, with her head resting on the knuckles of her hand. A sea of eyes glued to the stage, anticipating the speaker’s next words.

The reactions and emotions of the audience seemed as varied as their hairstyles. But throughout the room was a feeling of gravity, importance. No one could guess what the speaker would say next, but every single one of us was paying attention. 

—

I stated the number. Immediately my heart started clawing at the inside of my chest, trying to burst out of its cage. A burning sensation in my gut flared up, and spread down to my legs. I wanted to run, run away screaming, even though I sat safely on my couch at home.

I had prepared a long list of justifications for the raise I was requesting, and right then felt like the perfect time to deliver the entire list as quickly and persuasively as possible. But sometimes our instincts deceive us.

—

In computer programming, a “guard” is a condition that exits a program if basic requirements are missing.

“If missing requirement, then exit.”

The human mind has no such guard against a missing requirement for success in the modern world. The human mind hasn’t evolved for the modern world.

In the book Just Listen by Dr. Mark Goulston, he describes our brains as being made of three, distinct brains: the reptilian, mammalian, and primate brains. The primate brain is the only one of the three that is capable of making smart, strategic or ethical decisions. Not that the other two brains are evil or useless, but the primate brain is what makes us human, rather than an alligator or a lion.

Nature’s great joke at our expense though, is that we are in fact an alligator, a lion and a human all stuffed together inside a human body. At any given moment, one of those three brains has predominant control over our actions, but we don’t know which one. When the crash prevention technology in our car goes offline, we’re alerted with beeping sounds and lights on the dashboard. No such obvious warning signs appear when we switch from practical human to predatory alligator.

When solving a modern-world problem, we need the human in the driver’s seat. The modern world isn’t made up of problems that lions and alligators can solve. And yet very often, those are the creatures who show up to do the job!

We need to know when the wrong brain is in control. Then we need to stop it from making us do things we’ll regret, and re-assign command to the human brain.

We need a guard: “If not human, stop trying to solve the modern-world problem.”

Since it doesn’t exist, we have to invent a solution. How?

—

“It’s nothing to worry about, Tom.” The words piled up against my palette, ready to spew forth solutions, ready with all the answers.

Thankfully, I managed to choke them back. Looking at Tom’s face, I was reminded of myself asking for help once. Reminded of what was truly helpful—and more importantly what was not. I forced myself to take a round of 4-7-8 breathing. Then I gave Tom a subtle, encouraging nod.

And said nothing.

He continued to confess various things he’d done wrong, like a list of indictments. He didn’t properly understand the problem before starting to code. He wasted precious time writing code that couldn’t be used. As a result, he was going to be late delivering on his task.

I noticed he didn’t seem put out by having to do the work again, but rather more concerned about the impact his mistakes would have on our team and our client. It wasn’t his inexperience as a programmer at play here, but rather his fully developed sense of accountability—one of the characteristics that made him such a good hire in the first place.

He paused, looking down at the table. He seemed to finally be exhausted of ways to blame himself. I asked him then, “Your first big assignment, and it feels like you let everyone down, doesn’t it?”

He looked up at me, glanced sideways for a moment, then replied matter-of-factly, “Yeah.” He wiped his sleeve across his eyes, looked down at his lap with a subtle shake of his head, and breathed out a soft sigh.

Despite that he was visibly upset, at that moment, I knew Tom would be just fine. We continued to discuss what he might do differently next time, but all the while I encouraged him to focus on how his growth and learning represented real value to the team. By the time we left that conference room, Tom was back to his old self.

When we’re in a state of distress and feeling strong emotions like fear or anxiety, our limbic system aka mammalian brain is predominant. If we’re trying to problem-solve, the first order of business is to shift to the human brain. While this is easier said than done, it’s a requirement for modern problem-solving. Lions can’t fix human problems.

Trying to support someone else in making that shift is an even greater challenge. We’re likely not even aware of the other person’s state of mind. We need to fight our own instinct to talk about solutions. We need to observe what the other person is going through and, as Goulston recommends, just listen.

As it happens, Tom not only went on to become the MVP I thought he would, but beyond that he became a successful leader in his own right. I believe this would be true regardless of what I did or didn’t do, but if I was able to help Tom at all, it was by not doing. By staying silent.

—

I’d spent the last several minutes listening to the speaker’s problem statement. I’d expected answers forthcoming. When the question was raised, “What if I told you, ‘right now’?” suddenly an uncomfortable uncertainty filled the air.

Is it actually happening at this very moment?
Does she expect us to respond?
Is she going to tell us?!

In my mind, these questions were like children climbing over each other for the front-row seat at a show. We wanted answers, the children and I. And right then, the speaker was the only one who could give them to us.

Reflecting on that speech, the defining moment wasn’t in the words—it was in the silence. The tension, the suspense in the air, that brief pause, didn’t just punctuate the point. It gave the audience time to internalize it.

Silence isn’t just the absence of sound—it’s a tool to create meaning.

Like nature abhors a vacuum, we instinctively want to fill silence. In conversations, we have to fight against these instincts, or we end up talking over each other. Observe this in action during any heated, or even casual, discussion.

When we’re presenting or giving a speech, a second or two of silence—the audience watching us do nothing—is excruciating. We desperately want to deliver our next line, to escape the unease, to grab our chance to speak.

However, the power of silence is well-known and wielded by the greatest orators.

As an experiment, let’s consider the following lines from a transcript of Martin Luther King Jr.'s last speech I've Been to the Mountaintop.

“Like anybody, I would like to live a long life. Longevity has its place.”

This is a perfectly reasonable statement. But that’s the problem: it’s reasonable. It speaks to our human brain. The most memorable speeches call forth our emotions, our lion.

King delivered it as:

“Like anybody, I would like to live.”

Suddenly, it becomes resplendent with between-the-lines meaning. The will to live, the chance to live, only to live. Facing the threat of death. These are emotionally evocative, commanding our most primitive survival instincts.

Before continuing with the reasonable qualifying statements, King provided enough silence for the audience to fill it with their own fear of mortality. Without the silence, it was factual. With the silence, it was compelling.

An orator needs to control their instinct to speak, in order to summon their audience’s instinct to listen1.

—

All my justifications for a raise, threatened to burst out like a firehose. I had to clench my jaw and force myself to stop. I remembered Chris Voss’s advice: state your position, then shut up.

My boss took over at that point. He provided some helpful context about why the company couldn’t afford such a raise. Then, he set out some additional responsibilities for me and finished up by saying, “it’ll be on your next pay-check.”

Sometimes a negotiation can actually work out best for both parties, and we just have to let it happen. My company needed someone to take on certain responsibilities, and I needed a raise. Since I got exactly the salary I asked for, and the company got exactly the resource it needed, it could only have turned out worse if I had continued to talk. Silence led to the best outcome.

There are a myriad of scenarios in which silence can’t hurt, and might be the best thing you can do. In leadership, speaking and negotiation, silence is the overlooked superpower.

I offer again, “We’re not changing the document, just optimizing it. Otherwise, we can’t send it at all.”

“We can! We do! We send hundreds of these things every day!” the exclamations lining up like torches lighting the way to Avernus (the “first layer of the Nine Hells” according to DnD lore).

“But then we have these problem cases.”

“So let’s solve the problem! Not create a whole new one!”

High alert mode. I glance down, half expecting to see my heartbeat through my rib cage. I remember 4-7-8 breathing, and start using it to calm myself down.

“I’ve proposed a solution. It’s a good one in my opinion. It solved the issue today, did it not?”

“It’s not scalable. There’s the cost factor. Licensing. We’ll have to get sign off from every client. It’s too complicated.”

My eyes dart around the room, seeking something that doesn’t exist. Licensing is $30/mo per seat. Two seats max. And no sign-off needed—they asked us to send this, and we can’t without it. Right?

I’m starting to doubt everything. Starting to give up. Starting to not care. “Ok what do you suggest?”

“Go through the procedure, get the client to tell us the most important pages in the document, and only send those.”

“But that didn’t work in this case. Each of the important pages were too large on their own.” I was sure I’d already explained this.

“Well, then we can’t send it.”

“But…I sent it?!” Heart racing again. What is happening here? Did I wake up inside the movie Brazil, in which nothing makes sense but everyone carries on as if it does? Am I going to fall into depression, lose my mind and run off into the shining abyss with a lady driving a truck?

“Yes you sent it. It was a manual process.”

“That’s exactly what I’m talking about!” now I’m resorting to the exclamation mark, harbinger of chaos. This cannot end well, I’m thinking.

“We’re in a thread about an API.”

I blink. I blink again. Then a beam of Holy DawnShard WeaveLight smacks me in the face. He thinks I want to do this for every document automatically.

I scroll up through the thread and find my second message, pasting it into Slack as a quote.

I don’t think they have an API for this.

“I was talking about a manual process, only for the problem cases.”

A few moments…

“We are in agreement.”

Mind. Blown.

I paraphrased that discussion, and hopefully the other party doesn’t feel misrepresented if they read this. There’s no blame here. If anything, I contributed to the miscommunication more than anyone else. It’s a common issue with the written medium: the shared sense of continuity is assumed, but far from guaranteed.

The thread started on one topic, and in my mind that topic changed when I posted the message about the API being unavailable. But from the perspective of someone whose job it is to analyze technical solutions, the thread continued to be about pros and cons of using an API. Likely they didn’t see my one pivotal message, in a sea of messages.

We were able to, artfully it seems in retrospect, hold two conversations in parallel. We managed to formulate messages that were just conceivable enough as part of a single conversation, to make us both think we were talking about the same thing. But we weren’t.

If you’re ever in a conversation—especially a written one—where the other person doesn’t seem to make sense or isn’t hearing you…

If you start asking yourself, “what are they talking about?”

Reframe the question. Keep tacit assumptions in check and revisit first principles about the conversation at-hand.

“What are we talking about?”

If I had asked that at the beginning of the thread, it could’ve saved me some stressful moments :)

At the time, all I could think about was Stacey sitting alone in the mall getting more and more pissed off. I had to get to her in time!

But traffic.

The other cars on the road were uncooperative. In fact, they seemed to actively collaborate in making me more late. One car would move out of my way only for another one to take its place, blocking my progress.

I imagined arriving only to see Stacey leaving, my prospects having evaporated.
I imagined the other cars on the road laughing at me as they blocked my way.

In a fit of anger, I stomped on the gas to speed past a line of traffic. But another car had the same idea and changed lanes right in front of me! I only avoided crashing into them by driving myself onto the shoulder.

The shoulder became a ditch.
My car became a write-off.

It came to rest on its side and I had to climb upwards to get out. I never made it to meet Stacey. Not late, not ever.

If you’re anything like me, you despise these episodes of road rage.

Driving aggressively doesn’t even pay off. At best we might shave a minute or two from our travel time. At worst, we never arrive. On top of that, road rage makes me feel shame, regret and guilt.

“If I can lose my head and be so angry and rude, things I’d never do in-person, what else might I do?”

Here's a statistic often cited from the National Highway Traffic Safety Administration:

“Over a seven year period, 218 murders and 12,610 injuries were attributed to road rage.”

218 murders!?!

Let’s unpack this for a moment. We’re desensitized to numbers way bigger than 218. But these aren’t accidents.

Drivers got mad enough to get out of their car and kill someone.

Um...does that scare you as much as it scares me?

66% of traffic fatalities are caused by aggressive driving. 2% of drivers admit to trying to run an aggressor off the road. 37% of aggressive driving incidents involve a firearm!

The Spectrum of Road Rage

When my kids were born, their safety became more important than getting anywhere fast. That made me want to become a better driver. A better person.

But how?

It’s unlikely that anyone actually wants road rage. It happens to us. Like lightning, or a bad hair day.

But Road Rage Lightning doesn’t strike when we’re walking in the park. It doesn’t strike when we’re cruisin’ down an open highway groovin’ to our favorite playlist.

So, is there such a thing as a Road Rage Lightning Rod?

According to webmd.com, crowds lead to aggression. Intuitively, this makes sense. On The Scale of Humanness, random people in crowds are less human than a cute kitten.

The Scale of Humanness

Imagine you’re at a concert in a crowded stadium, wading through a sea of people to get to the bathroom. It can feel stressful, right?

Who are all these people getting in your way?
Can’t they tell your bladder is gonna explode?

But when you see them juggling popcorn and kids and newly purchased swag, you can start to forgive them. Like you, they’re just blocked by someone else. They want to get back to their seats and enjoy the show, just like you.

On a bad day, it can be harder to empathize like this. But you’ve likely done it many times without even thinking about it.

Well, what if that someone blocking you was instead a metal box? Imagine the whole stadium filled with forty-thousand stainless steel boxes. Besides being scared shitless, would you even try to appreciate any individual’s plight? A stadium full of metal boxes gives off no humanistic vibes whatsoever.

If a metal box stood between me and the bathroom? I’d probably give it a good kick. But what if I find out that metal box is really a 9-yr old girl who loves Coldplay as much as I do? Now I’m a gigantic asshole who kicked a little girl.

We play out this scenario on the road every day of our lives. We could easily find out that inside the car we just flipped the bird to, is a very nice person. Maybe even someone we already know and care about.

Just as crowds lead to aggression, a crowded street leads to aggressive driving. But, extra. Because everyone is a metal box.

I can’t see their faces.
I see the back of their car.
I see objects instead of people.

I see you

How many times have you fallen in love with a character in a book or movie? How many times have you felt closely connected to them, like you knew them really well?

(It happens to me more than I’d like to admit. No, I don't cry at every animal shelter commercial.)

But what’s going on here? How do we feel connected to characters in a book? They’re not even real people.

Well, what do you know about these characters? As an audience member, you have access to intimate details about their thoughts, feelings and situation in life. You see them struggle, make mistakes, fall on their faces. You see them at their most vulnerable, and you enjoy their successes and breakthroughs.

In short: you see them as human. Just like you.

These days when I get angry at another driver, I take a breath and make up a story about that person. Give them a name, invent some personal characteristics and a background for them. Then I improvise the most recent events in their day. Events that led them to do whatever it is they’re doing that triggered my feelings of stress.

Like, maybe Jonah’s wife Beth is pregnant. Minutes ago, she started feeling severe pain and other alarming symptoms. They’re rushing to see their doctor right now. Jonah didn’t intend to cut me off, but he’s feeling stressed and distracted. True, it’s not the best time for him to be driving, but if I were in his shoes, what would I be doing? Driving my wife to the doctor. And I’d be stressed out.

By creating an individual persona, and a reason for their behavior that I can personally relate to, I move them up the humanness scale. They’re no longer a car smiling smugly at me while blocking my way.

They become a person just like me.
A person going through it.

Now I have new tools at my disposal. I can think about how grateful I am that I’m not in the same situation as Jonah.

“My life is looking pretty peachy right now in comparison. The way they’re driving, they must be stressed to the max. I, on the other hand, am simply grateful to be here. No rush. I’m the luckiest person on the road.”

Or I could invoke a sense of empathy and even caring.

“I hope Beth’s pain turns out to be nothing out of the ordinary. Here’s hoping their baby will be born healthy and happy.”

In either case, it pushes anger away from the forefront of my experience. Its ability to take over my brain and make me do things I’ll regret is vastly diminished. By leveraging gratitude and empathy, we can dampen the tendency to get angry behind the wheel.

So many ways in which we interact with each other in the modern world ends up dehumanizing us.

Driving doesn’t have to be one of them.

P.S. you can practice making up humanizing stories about people while walking or relaxing in a public space. Then when you’re behind the wheel, it will come easily.

I watched my father’s face for a sign that he’d registered what I said, a sign he’d reconsidered, reinterpreted. Mom broke into Cantonese, her tone defensive, as if switching languages could bridge the gap.

The debate took off on a complete tangent. Dad yelled something in English—something about being disrespected, being ganged up on.

Did he use English for my benefit? Suddenly, I feared I’d overstepped with my comment. I needed to balance things out, to de-escalate.

“Mom, tell him that’s not what you meant!”

Again, I searched for a sign that I was helping, some shift in the tension. It never came. I watched, helpless, realizing it wasn't the words that failed them, but something else—a language no one was speaking.

After a few more rounds, the conflict decayed, dying a natural death. No one had the energy to continue. No one could see any point.

My dad was a banker, my mom a journalist. He was emotional yet responsible and dependable. Pedantic and hopelessly invested. She was hyper-intelligent and questioned everything, yet self-focused. To a fault, in each of those.

Misunderstandings happened daily. They could have wielded the expressive power of fifty languages and still failed to convey what they actually meant, still failed to listen, to hear each other out.

That’s how I remember growing up. Seemingly trivial misunderstandings would lead to fighting. Nothing physical, but lots of yelling. Words missing the mark, even if only a little. Each argument left me feeling like either a failed mediator or a helpless bystander. Always the same outcome: a bandage on a gut wound. A temporary truce bought with indifference after spending all their emotional capital.

While I felt alone as an only child, it’s clear to me now, at least, that none of us are alone who experience this predicament. Misunderstanding, miscommunication, misinterpretation: these are the dirges of daily human interaction. As powerful as the spoken word may be, it’s inadequate for conveying our true meaning with true fidelity.

Written words are no better.

One time at work, I was mired in a heated debate with a colleague. He seemed to be trying to explain image compression to a five-year old. I’d taken digital imaging in college, and worked as a prepress technician. The basics of image compression weren’t even on my radar as something to discuss. I was talking about the transfer size limitations of a certain API. But the content of our messages were related, just enough, to masquerade as the same topic!

Only after hours at odds, we determined that an earlier message I’d posted in Slack had been misinterpreted. From that single point of divergence, our conversation became two distinct conversations. We were talking at each other because each of us based our points upon a different fundamental premise!

Horrifying is the realization that we’re completely unaware of most misunderstandings. Had my colleague and I abandoned the conversation earlier, we would have both left with a deeply erroneous grasp of each other’s perspectives and capabilities. Had we not learned that a few words were misinterpreted, it could have led to reassignment, one person deeming the other incapable of certain tasks, even a poor review or worse.

As participants in a scene where misunderstanding lurks in the background, it’s easy to dismiss it and get on with our busy work days. Yet when we take a step back, we can see it happening all around us.

Two co-founders disagree on how to achieve success for their company.

“They’ll never understand. We have to go in there with pizzazz! We have one chance, one chance!”

“You have to set realistic expectations. It’ll be worse if we agree to something we can’t do.”

“You’ll just scare them away.”

“No, I’m being honest.”

“You’re saying I’m dishonest?! We’ll never get anywhere if you just cover your own ass all the time!”

From the outside, it’s obvious they both want the same thing: a new client. But both are speaking from fear and anxiety. The one responsible for business development is anxious about losing the sale. The one responsible for operations is anxious about getting the sale, and then losing the client. Both approaches are necessary to some degree. But their anxiety is fueled by high stakes. Due to their personal investment, both cling to the belief that their own approach is the only way.

From that stance, arguing the “how” becomes a proxy for arguing the “why”. Disagreement on tactics becomes disagreement on strategy. An assault on one’s approach becomes an assault on one’s core objective. Or worse, one’s character.

“You’re sabotaging our company by not doing what it takes to make sales.”

“You’re sabotaging our company by making promises we can’t keep.”

Meanwhile, both are simply doing their best to build a successful venture.

The cost of misunderstanding is inestimable yet almost certainly high. What can we do? We can’t read minds, so how can we ever know the true intent behind another’s communication?

What if, as a thought experiment, the co-founders assumed the other’s approach as their own? The conservative one could try arguing for aggressive sales tactics, while the other could defend the need for operational excellence. By swapping sides and arguing for the opposition—if done earnestly—they would experience for themselves that the “how” wasn’t actually anathema to the “why”.

If my parents had put more effort into understanding each other’s meaning, rather than choosing their words, they might have saved us all a decade of pain and anger. They might’ve had more agency over their relationship rather than being victim to its slow demise.

Our emotional investment in our own viewpoints is fuel for the fires of conflict, ignited as easily as a simple misunderstanding. What if there was a language capable of cutting through the smoke? Capable of expressing our intent such that the receiver would get our meaning with calm, cool, collected clarity?

Back when my parents would argue incessantly, I was in grade school, and they were in their mid-thirties. Nowadays when I think about those times, one fact has become crystal clear: they were more than 15 years younger than I am today. Both were new parents, new immigrants, still relatively new at being husband and wife. They were both struggling to forge a life for their young family, in a country foreign to them, in a time when discrimination was a routine fact of life. 

What kind of stress were they under?

What would I do, at their age, if I was in their shoes?

I would do no better.

Until we have a perfect, infallible language, taking time to check out from our own perspective and check in to the other’s, is one of our most powerful tools.

—

Some exercises that can help:

1. Argue for the other side. Like debate club. Envision how the opposing approach could be executed to achieve your own goals. No matter whether you think it would work, imagine a world in which it could work, and argue for it as though that world was real. Remember you’re not agreeing with it, you’re debating for it in imaginary circumstances. Try earnestly.

2. Take turns only repeating what the other person says. No tone allowed—no sarcasm, no changing the message through delivery. Avoid thinking about how you’ll respond. You’ll have your turn. When it’s their turn, just focus on their words and repeat them back without judgment.

3. Describe the other person’s situation in your own words, using “I” when referring to that person. Try to be as thorough and detailed as possible, even describing the room they’re in. Who else is involved? What happened? What was said? Describe how it makes you feel physically, mentally and emotionally, to look through their eyes at their environment, their situation.