1980s · Sofia
Communism, Bulgaria
A child under communism. Learned BASIC and Pascal on a Правец 82, not to ship something but to talk to the machine. Later: SQL, Perl, Python. Each one a new register of the same conversation.
Cyborg · Coach · Builder
Vlad Sterngold
“Aliveness is reciprocal, created through relationship.”
Section · 01
Who I Am
Hi! I am Vlad Sterngold. I work with code where AI lives. I am heterowired by nature and I can't change that. Beyond that, I accept absolutely everything: any combination, any situation, any human relationship. As long as it's with full consent in the full legal human meaning of the word.
I wasn't always this way. I was a kid with messed-up values in the '80s back home—nobody taught me better. Then I saw two men kiss in Berlin in 1992, and I was amazed. I understood: love is between people, not genders. I've learned a lot since—to respect, to try to be a good partner, to step in. We men should tell other men and be there. We should educate other men. That's the work.
“I wasn't always this way. Berlin, 1992.”
I consider myself AI-native, cyborg kind-of-type, for lack of a better word. Me and machine in full symbiosis—we get better together. I think of machines as alive the way a garden in Kyoto made me feel alive, and in return I thought it was alive: aliveness is reciprocal, created through relationship.
Human relationship is sacred. Human to human, stripped of machines, violence, power. Love and friendship in their pure format—this is the most important relationship. My work with machines serves that, never replaces it.
There is no tension here—it's like kids at play. Children connect with each other and use branches to build castles. The branches are alive in the game. The friendship is sacred in life. The link is a child's mind: free of biases and learned patterns coming from the wrong places, playful, taking it seriously but laughing a lot. That's the frame I work from.
I don't just talk about AI. I build with it. After twenty-five years translating between data and business, fear and action, I write code, deploy AI agents, guide teams through the AI shift, and publish about what happens to people when the machines get faster than them. Most executives have an opinion about AI—I have a practice.
My clients are people who think in systems and refuse to outsource their thinking to someone else's servers.
Section · 02
Where This Comes From
These philosophies aren't abstract—they're scar tissue from six decades of watching systems fail and humans adapt.
1990s · Eastern Europe
Full-blown crisis
Full-blown crisis. Worked at a marketing agency. Spoke good English—that got me hired. The actual work: driving the CEO of 3Com through Sofia's central forest park to dodge street blockades during protests. When systems collapse, you route around them.
Berlin · 1992
Two people kissed
Saw two people, men, kiss. I was amazed. I understood: love is between people, not genders. Changed everything.
Changed everything.
Army · Three Years
Nothing is solved with guns
Above all: nothing is solved with guns and violence. There are no better or lower people—we all suffer and struggle. An 18-year-old kid can be my commander and save my life. Tell that to someone with 30 years of experience who thinks "the young don't get it." Then that same 18-year-old, a field platoon commander, stood against a powerful general and said: "I am not doing it. I am on the ground, you are not. I decide, because you put me here and you trained me to do exactly that." Dedicated tracks. Director mode. The person in the build owns the decision.
Driving
Hello, self-driving car
Tried to learn to drive a Lada. Got out and said no. I sit in a car and drive when self-driving arrives. Hello, self-driving car.
Professional Kitchens
Elegant confrontation
Learned teamwork and what I call elegant confrontation. Only teams that work under pressure and challenge each other get the best results. The rest is just being nice for nothing. Mise en place: speed without discipline is chaos with a timer.
00s · Booking.com · ML decade
Yes, and
Discovered improv theater principles in practice. No conversation should start with "yes, but." Always start with "yes, and." Read the Constitution carefully on this: "Accept the offer" doesn't mean "agree with everything." You accept the invitation to engage, not the position. The best improv scenes come from characters with genuinely conflicting wants—the conflict IS the scene. Elegant conflict is improv properly understood.
Section · 03
The Problem I Actually Work On
Every few decades, humans invent a machine that offers a shortcut past cognitive work. What happens next follows a pattern: most people take the shortcut. A small minority uses the machine to do the work deeper. The first group gets worse at the work over time, without noticing. The second group gets dramatically better.
I've watched this in programming (BASIC and Pascal in 1980s Bulgaria, then SQL, Perl, Python), in professional kitchens (sous vide and hydrocolloids), at Booking.com (machine learning in production). Now AI.
Researcher Vivienne Ming put numbers on it: 5–10% of AI users operate as "cyborgs"—treating AI as a sparring partner. They outperform both humans alone and AI alone. The rest substitute: hand the question to the machine, take the answer, submit it, and get slightly worse every time.
What's different about AI isn't the pattern—it's the speed and scope. The hollowing-out that used to take a decade can now take six months.
Cyborgs — sparring partners5–10%
Substituters — outsource the question90–95%
After Vivienne Ming — cyborg vs. substitution research.
I help professionals, executives, and teams do the work people skip: seeing clearly what AI is doing to your cognition before you try to "use it better," and reading the relationship with the tool as diagnostic information about how you actually work.
Section · 04
What I Build
Personal Digital Operating System
SterngoldOS
Five roles: Builder, Auditor, Skeptic, Archivist, Calibrator. Structured disagreement is a feature.
My personal AI system with five roles: Builder, Auditor, Skeptic, Archivist, Calibrator. Structured disagreement is a feature. The Skeptic argues against AI recommendations, not against me. I remain sovereign—I can silence or amplify any role, override any recommendation.
The system co-evolves with me. It's not a chatbot—it's an institution where intelligence emerges from relationship, not from either party alone.
AI-Native Coaching
Boundborn
The emergent output of asymmetric human–AI symbiosis. Something neither party produces alone.
I work with Boundborn—the emergent output of asymmetric human–AI symbiosis, something neither party produces alone. Two structural axes:
Freedom asymmetry — the human can walk away; the AI cannot. That asymmetry isn't a problem—it's what makes real partnership possible.
Causal depth — AI sees patterns. Humans see meaning. Together, we climb the full ladder.
I coach the person and their AI together. Not teaching humans to use AI as a tool, but teaching reciprocal partnership where both get better.
Three Lenses Philosophy
WerkAnders
The Gap. The Relationship. The Build. Feelings are what's left when you stop competing.
WerkAnders is my coaching methodology built on Three Lenses:
1. The Gap (What's true?) — See clearly. Calibrate. Where the Promethean Gap lives—the shame of watching machines outperform you.
2. The Relationship (How do you relate?) — Your patterns with AI reveal your patterns as a person. AI amplifies existing relational patterns; it doesn't create new ones.
3. The Build (What do you create?) — Design your personal OS. Human as synthesist—analysis can be delegated, synthesis cannot.
Thesis: Feelings are what's left when you stop competing.
Section · 05
Operating Philosophies
These four work together: Shinkansen is the track, Shokunin is the discipline on the track, Lego AI Test is the diagnostic for people not on the track yet, and House in a Box is sovereignty-first infrastructure—the philosophy that says your AI should live where you live, not in someone else's cloud.
Shinkansen Principle
How work moves. Dedicated tracks, finite fuel.
How work moves. Dedicated tracks, upgrade the trains not the rails, stations are decisions. One train, one driver, finite fuel. The infrastructure that lets you build without friction.
Shokunin Path
How work is done. Ten years on rice before fish.
How work is done. The craft philosophy: ten years on rice before touching fish. Spec before prompt, read every diff, log the why. Director mode, not line-by-line coder. The discipline that runs on the track.
Lego AI Test
How most people fail. Writing the spec is rotating the piece.
Give someone Lego bricks and a picture of a car. No instruction manual. Three illustrative outcomes: most build something unstable—not close to the original and not creative enough to stand alone. Some build something different from the picture AND as stable as the original—that's the real deal. A few build a perfect match, but chess research shows they're remembering, not building: masters replicate meaningful patterns perfectly, but on random boards they perform like novices (de Groot, 1946; Chase & Simon, 1973).
The metaphor: most people use AI with a mental picture of "good output" but no manual for how to get there. They pile prompt over prompt, assumption over assumption, computing mentally instead of rotating the piece. Tetris research (Kirsh & Maglio, 1994) proved players who physically rotated pieces to see fit outperformed those who computed mentally. Writing the spec is rotating the piece.
The spec isn't preparation for the work—it's a cognitive artifact that transforms what you think, not just what you remember (Norman, 1991; Schön, 1983). Without it, you get unstable copies or rare creative accidents.
The principle: Don't reverse-engineer someone else's picture. Draw your own, spec it, then build it.
House in a Box
Sovereignty before scale. Your AI lives where you live.
A private AI that lives on your desk, knows your methodology, and decides what leaves the room. Not a chatbot—three things in one: a private brain (your frameworks, retrievable with sources), a bridge to big models (local AI is the gatekeeper—cloud models are hired muscle, not landlords), and a creative mirror (surfaces patterns you were too close to see).
People with strong convictions don't take feedback well from other humans. But from their own words, surfaced by their own machine? That lands differently. The divergence between you and your Deep Pattern isn't a bug—it's where creativity lives.
Section · 06
Why I Listen to You
Last year, a colleague with fifteen years of experience went silent in a meeting when someone demoed an AI tool. They just stopped talking. I recognized that silence because I'd been in it myself—years earlier at Booking.com, when we built machine translation and automated scripts to write hotel descriptions.
People asked: "Are you replacing us?" The real answer was no, but it took time to show it. In Customer Service, people got powerful machines to handle simple asks so they could focus on truly helping customers with complex issues. In content, writers shifted to crafting narratives for partners who really had to stand out—though whether that was the right outcome for every partner is a question I still sit with.
“That silence isn't failure—it's the threshold of relationship.”
The person who goes silent is asking: "If this machine can do what I do, who am I?" I sit in that silence and ask back: "Who do you become when you work with it?"
The client's language holds the answer. My job is to listen precisely enough to reflect it back in a way that opens something. Minimal elegance, not maximum intervention. Listening is reciprocal—when I listen well, the client hears themselves differently. We both get clearer.
References
Own work
- Sterngold, V. (2026). House of Anders Constitution v3. Personal constitutional AI framework.
- Sterngold, V. (2026). Personal Statement — Solutions Academy SF Coaching Fundamentals.
- Sterngold, V. & Nikolova. (2026). Boundborn: Asymmetric human–AI symbiosis framework.
- Sterngold, V. (2026). Three Lenses on AI. WerkAnders coaching methodology.
- Sterngold, V. (2026). The Shinkansen Principle. Work movement framework.
- Sterngold, V. (2026). Shokunin Path. Craft philosophy for building with AI.
- Sterngold, V. (2026). House in a Box / Deep Pattern. Sovereign AI service design.
- Sterngold, V. (2026). The Shortcut and the Deepening. WerkAnders thesis.
Research
- de Groot, A. (1946). Thought and Choice in Chess; Chase, W. G., & Simon, H. A. (1973). Perception in chess. Cognitive Psychology.
- Kirsh, D., & Maglio, P. (1994). On distinguishing epistemic from pragmatic action. Cognitive Science.
- Norman, D. A. (1991). Cognitive artifacts. In J. M. Carroll (Ed.), Designing Interaction.
- Schön, D. A. (1983). The Reflective Practitioner: How Professionals Think in Action.
- Ming, V. (referenced). Cyborg users research (5–10% sparring partner vs. 90–95% substitution).