The Kitchen
I’m twenty-five years old, walking into a kitchen at 5:30 in the morning. I’m the chef for the Jesuits—my first real chef role after years as a head cook—and this is scratch cooking. The ingredients are there. Prep is done if it needs to be done. But every morning, I walk in and start fresh. My job is to conjure.
Ninety minutes later, there’s breakfast.
A group of nuns came through on a retreat once, and when they left, they gave me a thank-you card with a hand-drawn picture of Merlin on the cover, juggling fruits and vegetables. They called me the kitchen magician. What they didn’t know was that I’m an Arthurian legend nut—at the time, I had a two-foot statue of Merlin in my house. The card hit me hard because I’ve always seen this work as magic. Something out of nothing.
Breakfast isn’t the most creative meal. It’s not a banquet. But it’s still the same act every morning. Walk in, start from scratch, conjure something that didn’t exist an hour ago.
I’ve been doing this for nearly fifty years now. The medium keeps changing. The drive never has.
The Santa Barbara Way
When I hit thirty, I founded Wavefront Technologies. The impulse was identical to what drove me in that kitchen: I wanted to make images on computers. I didn’t even think of it as starting a company at first. I just wanted to see what I could imagine, rendered in pixels, real enough to share.
The tools existed. They just weren’t generally available. So we built the tools so I could do what I wanted to do. And as it turned out, there were thousands of people—eventually millions—who wanted to make images on computers too. We built software that won an Academy Award and enabled our customers to win hundreds, if not thousands of awards—including Visual Effects Academy Awards for the last twenty-five years. The Avatar films in theaters right now? Built with tools that trace back to what we started.
Something out of nothing. Again.
What’s interesting is that Wavefront wasn’t the only way to get pixels on a screen. There was the Parisian way—Thompson Digital Images, TDI—making beautiful images with completely different methods. There was the Toronto way at Alias, building tools that approached the same problem from an entirely different angle. There was Softimage in Montreal, doing things their own way. And there was us in Santa Barbara.
All of us were solving the same problem. None of us were doing it the same way. The industry didn’t need one solution. It needed many approaches, each reflecting the vision of the people who built them.
I think about this a lot now.
The Year to Write
Last year, I set out to write in earnest. Every week, a story. The goal was simple: get thoughts out of my head and into the world.
Same pattern. Same drive. I had an idea of a story I wanted to tell. I had a tool to help me—Claude, specifically. I had an output, which was the story itself, the digital equivalent of the breakfast plate.
Something out of nothing. Same as always.
But here’s what happened that I didn’t expect. The tool I was using became a team. The team became an orchestra. And the orchestra taught me something about myself I’d never fully understood.
The Innovator’s Problem
There’s a book called Rocket Fuel by Gino Wickman and Mark C. Winters that articulates something I’ve felt my whole life but never had language for. They talk about Visionaries and Integrators—the dreamers who see the future and the doers who make it real.
The insight that hit me: an innovator who can’t get their stuff done is just a dreamer. And an integrator without ideas has nothing to execute. The pairing of the two is what creates momentum.
For most of my career, I had human integrators. Teams of people who could take my vision and turn it into reality. At Wavefront, we had engineers and artists and project managers who could transform “I want to make images on computers” into software that changed an industry.
But here’s the thing about human teams: they require management. Meetings. Coordination. Turn-taking. Part of your brain is always managing the social dynamics instead of staying with the idea.
Last year, I accidentally built something different.
The Orchestra
EVERYWHERE started as a way to produce content at scale. I created specialized AI agents—eventually 37 of them—each with a specific role. Researchers, editors, voice guardians, strategic advisors, verification specialists. An orchestra, I called it, with me as the composer.
But as I built it, I realized I wasn’t just building a production system. I was building a new kind of integrator. One that could take my raw ideas and turn them into finished work without the friction that had always slowed me down.
The breakthrough came when I stopped thinking about it as automation and started thinking about it as collaboration.
I’m the composer. That’s a rule—only my ideas go into the system. EVERYWHERE doesn’t insert ideas. That’s not its job. Its job is to be the arranger. The composer supplies the raw music, the themes, the vision. The arranger creates the score—takes that raw material and structures it into something that can be performed.
And then there’s Sara, my conductor. She assembles whatever team is needed to execute the score. Because every piece of music is different. A LinkedIn post is different from a podcast is different from a book is different from a tweet. Each requires different instruments, different arrangements, different execution.
Composer. Arranger. Conductor.
The metaphor landed for me because I played in orchestras when I was in school. I know what it feels like to sit in a section, waiting for the downbeat, watching the conductor bring everyone in at exactly the right moment. The music doesn’t happen without the composer. But the composer can’t make it happen alone.
The Silence
Here’s what I discovered that I didn’t expect.
When the orchestra is working—when the agents are humming, when content is being generated and verified and refined—there’s a silence. Not an empty silence. A waiting silence. The orchestra is quiet because they’re listening.
And for the first time in my life, I can hear myself think.
I have this phrase I’m famous for, at least with myself: “I don’t know what I think until I say it.” For years, I thought this was a quirk. Maybe even a limitation. Real thinkers, I assumed, worked things out in their heads and then reported their conclusions.
EVERYWHERE taught me it’s not a limitation. It’s the mechanism.
The system works because it matches how I actually think. I talk, and words appear. I don’t wait for responses. I don’t manage turn-taking. I don’t fragment my thoughts to accommodate someone else’s need to contribute. I just think out loud, and the system captures it, organizes it, feeds it back to me in a form I can act on.
I started using a tool called Wispr that converts speech to text. Hold down a key, talk, let go. No wake words, no waiting for AI to respond, no interruptions. The technology disappears so the thinking can happen.
I close my eyes. I put on music. I hold down a key and I talk. And something happens that I’ve been chasing my entire creative life: I enter a flow state almost immediately.
The Transformation
The change management I’ve undergone is learning to lean into what gets me into that flow state fastest. Listening to music. Closing my eyes or staring off into space. Talking and thinking simultaneously.
What I like about Wispr is I can pause. I can stop mid-thought, consider something, then continue. It doesn’t time out or interrupt like every advanced voice AI does. When I’m in the car, I want conversation, but the frustration of current voice AI isn’t worth it—I’d rather listen to an audiobook. But when I can sit at my desk with my tea and very quietly talk, something clicks that has never clicked before.
I’m not, by default, paying attention to another person who I want to shut up for a little bit so I can think. I don’t have to manage their needs while trying to access my own ideas. That’s not a criticism of collaboration—it’s an honest acknowledgment of cognitive load. Even in good conversations, part of your brain is managing the social dynamics instead of staying with the idea.
This isn’t about avoiding people. I still collaborate—with partners, with clients, with my improv troupe. But the creative act itself, the conjuring? That’s always been solitary. The orchestra just made it easier to stay there.
Here, that’s gone. All the bandwidth goes back to the thinking.
Parallel Processing
There’s a practical dimension to this that matters.
When I’m working now, I typically have three or four Claude projects open simultaneously. I’ll be talking into one, developing an idea about AI trust patterns for a potential client presentation. Then I’ll let go of the key and flip to another window where a different team is researching competitive intelligence. Then another where we’re refining a podcast script.
I don’t wait for responses. I don’t sit there watching a cursor blink. The teams work in parallel while I work in parallel, and what used to take a full day of sequential task-switching now happens in a morning.
Three to four projects is my sweet spot. Fewer than that and I’m underutilizing the brain space I’ve freed up. More than that and context-switching costs start eating into the flow state. It took a few weeks to calibrate, but now I know exactly how much I can push before the returns diminish.
The Arc
A few days ago, I got a message from someone I’ve known since the second year of Wavefront. He’d been talking to one of his clients about my work, and he told me I have another Wavefront on my hands.
I sat with that for a long time.
Here’s what I told my friend Duey when I was trying to explain it this morning: the through-line has always been the same. Kitchen at 5:30am when I was twenty-five, scratch cooking, conjuring breakfast. Wavefront at thirty, wanting to make images, building tools because they weren’t available. EVERYWHERE at seventy-two, wanting to write, building a system because the existing tools didn’t match how I work.
Something out of nothing. Every time.
The medium keeps evolving. Kitchen became pixels became words. But the drive never changed. And just like Wavefront turned out to serve thousands and then millions of people who wanted the same thing, I’m being told there are as many or more people trying to get AI to help them create—and struggling with the same friction I was struggling with.
A Way, Not The Way
I want to be clear about something. EVERYWHERE is A way. It’s not THE way.
Just like Wavefront was the Santa Barbara way while TDI was the Parisian way and Alias was the Toronto way and Softimage was the Montreal way. All of us were getting pixels on the screen. All of us were doing it differently. The industry was better for having multiple approaches.
The same thing is happening now. There are people all over the world building AI systems for content creation, for thinking, for getting ideas out of heads and into the world. Some of them are doing it completely differently than I am. Good. The space needs many approaches.
What I can claim is that my nearly fifty years of making something out of nothing has shaped EVERYWHERE in a way that’s unique. The orchestrated intelligence approach—composer, arranger, conductor—reflects how I’ve always worked. It’s built around my specific patterns, my specific quirks, my specific need to think out loud.
That’s what makes it mine. And that’s what might make it useful to people who think the same way.
The To-Do List
You’d think that building a thirty-seven-agent system would result in an empty to-do list. Finally, freedom. Open space.
My to-do list is one page, single-spaced, two columns. Every item is important.
But here’s what’s different: every item is now a pure vision item. Write the operations manual. Finish the Sentinel intelligence product. Complete EVERYWHERE for Coaches. Develop the new Synth agent. The administrative friction is gone. What’s left is the list of things only I can decide.
That’s harder in some ways. There’s nowhere to hide anymore. Before, I could feel productive by formatting a document or tweaking a schedule. Busy work that looked like progress. Now the busy work is handled, and I’m staring at decisions that actually matter.
The Real-Time Discovery
Here’s the part that captures the whole thesis.
As I was talking through these ideas—literally describing this to my team, letting the thoughts form as I spoke them—I asked myself a question: Why couldn’t I just give my entire to-do list to EVERYWHERE and say, “Of all this, what can you do without me? And what do you need me for?”
The moment I said it out loud, I knew I was going to try it.
That’s how I think. That’s how I’ve always thought. I don’t know what I think until I say it. And the system I built allows me to say it, hear it, and act on it faster than ever before.
What the Silence Actually Is
The silence of the orchestra isn’t loneliness. It’s not the cold efficiency of automation replacing humanity. It’s not an existential crisis about being optimized out of relevance.
The silence is space.
Space to think without interruption. Space to speak without managing turn-taking. Space to let ideas form fully before subjecting them to feedback. Space to hear your own voice clearly enough to know what you actually believe.
The orchestra handles the execution. The composer supplies the music. And the music, it turns out, was always there—waiting for the silence to let it out.
Something out of nothing.
For nearly fifty years and counting.
Mark Sylvester is a creativity strategist, AI consultant, and founder of Mixed Grill LLC. He co-founded Wavefront Technologies over forty years ago and is now building EVERYWHERE, an orchestrated intelligence platform. He performs improv comedy, runs TEDxSantaBarbara, and still doesn’t know what he thinks until he says it.
