Q1 '26

Building a new machine

Q1 is the quarter of beginnings. New year, new resolutions, goals set and launches made. It’s the quarter where you plant things on purpose, where the calendar is still clean enough to be ambitious and the year is still young enough to make a significant start on things. Some years, a few of those seeds take root and the rest quietly die on the vine by March. This year was different. This year, nearly everything I planted in January was still growing by the end of March, and some of it had already started producing beautiful results. Q1 of 2026 was the quarter I invested, deliberately, seriously, with real time and real money, into the things that come next.

I built an agentic AI system I expect to use for years. I watched my COO take his first steps toward building generational wealth for his family. I walked my daughter through a city that rearranged her sense of the world. I followed my son into his interests and found real beauty there. I delivered a keynote about irreversible decisions and answered my first federal call for public comment. I watched some of our investments payout and immediately reinvested them into more growth vehicles. I counted 34 nights away from home, down from my trendline , and I felt like each one was spent well. The quarter had a theme, and the theme was planting things I intend to harvest for decades.

Where I Went

Scottsdale:

n mid-January I flew to Scottsdale for my second session with the Optimus Mastermind, which I joined in Q3 of last year after Brad Hart invited me to Phoenix. That first session encourage me to build skills around making and shipping technology using AI. This one changed how I live with it.

Optimus is not a conference. There are no panels. There is no keynote stage. Nobody is wandering the halls trying to sell you a SaaS product. You show up with your most important workflows and a roomful of people who are actually deploying agentic systems in their businesses help you automate them. I brought real problem, scheduling inefficiencies, content production bottlenecks, financial reporting that was eating hours every week, and left with architectures I started building on the flight home. The working sessions ran three days. The conversations over dinner ran longer. I ate at Bourbon & Bones in Old Town Scottsdale with some new friends, which was the best meal of the quarter, and given that Paris was also on the itinerary that is not a small claim.

The real value, as always with these things, is the people. There is a widening gap between folks who are actually building with AI and folks who are debating which LLM to use. The Optimus room is the former and every person at that table ships. That changes the quality of the conversation dramatically. You skip the “is AI real?” part and get straight to “here’s what broke, here’s what I tried, here’s what worked.” I cannot overstate how much more useful that is.

Atlanta — Tommy Breedlove

Tommy Breedlove is one of my closest friends in the world, and if you’ve been reading this newsletter for any length of time you’ve heard me talk about him. When he asked me to speak at his event, Legendary Live, in Atlanta, I didn’t think about it for longer than the time it took to say yes.

I delivered Rubicon, a talk I’ve been developing about the moments in life and in business where you cross a line you don’t come back from. The word comes from Caesar’s crossing of the Rubicon River in 49 BC, which was the moment he committed to bringing meaningful change to Rome. Every supreme decision has a Rubicon embedded in it: the choice to leave a company, to start one, to end a relationship, to begin one, to tell someone the truth they didn’t ask to hear. The talk is about recognizing those moments while you’re in them rather than only seeing them six months later in the rearview mirror. Most people realize they’ve crossed their Rubicon long after they did it. The goal is to close that gap, because a conscious crossing is one you can own and an unconscious one owns you.

Tommy’s audience is his tribe: driven, thoughtful people who have already succeeded by most external measures and are now trying to figure out what success looks like when the scoreboard stops being interesting. They were generous and engaged and I took as much from the room as I gave to it. Tommy opened, I delivered the keynote, and we spent the rest of the conference doing what we always do when we’re together—trying to outgive each other.

Atlanta — Manga with Ret

My son Ret got into manga this quarter, and I want to talk about that for a minute, because I think it matters more than it looks like it matters.

Manga is the Japanese tradition of serialized graphic storytelling. In America we call them comic books and treat them as disposable entertainment. In Japan they are a literary tradition with serious writers producing serious work that sells millions of copies and grapples with questions about identity, morality, sacrifice, loyalty, and what it means to live a good life in a world that does not make that easy. The best manga series are not superhero stories. They are moral philosophy delivered through narrative art, and if that sounds like a stretch, you haven’t read the right ones yet.

A friend and mentor told me years ago, long before I had children: “You become interested in the things your kids are interested in.” He wasn’t talking about tolerating their hobbies with a patient smile. He meant that you follow them into whatever they love and you let it teach you something. I took that advice with Ret this quarter. We went to some of the specialty bookstores in Atlanta to hunt down the titles he wanted, and to show him I was genuinely interested and not just chaperoning, I asked him to pick out some titles for me.

I grew up on comic books and the lessons they taught me. Spider-Man taught me that “with great power comes great responsibility,” and the X-Men taught me that “to go far, go with,” and Iron Man taught me that your greatest power comes from where you hurt the most. These are the moral frameworks of my adult life, and what I remember most is not the whiz-bang action of the comics but the moral and emotional situations they presented for my own analysis. Should Wolverine kill the villain or let the justice system handle it? What does Spider-Man owe the city that hates him? Should you stop being smart when it starts pushing people away? These were the first serious ethical questions I ever encountered, delivered in panels instead of paragraphs, and they shaped how I think about right and wrong to this day. Manga is doing the same thing for my son’s generation, and in some cases does it better, because the Japanese tradition is less afraid of ambiguity, less committed to happy endings, and more willing to let the reader sit with discomfort.

Ret picked two series for me, both by the same author, Naoki Urasawa. I’ve started Pluto, which reimagines a classic Japanese storyline as a murder mystery set in a world where robots have achieved something close to consciousness. A robot detective investigates a string of killings targeting the world’s most advanced robots and the humans who championed their rights. The mystery is compelling, but the real subject is the emptiness of war, what it does to the victors, what it costs the occupied, and whether the machines we build to fight our wars inherit our grief along with our orders. It’s heavy and it’s beautiful.

I’ve also started Monster, Urasawa’s masterpiece, which is a psychological thriller about a Japanese neurosurgeon working in post-reunification Germany who saves the life of a young boy and discovers years later that the boy has grown into a serial killer of extraordinary intelligence and charm. The central question is one that has haunted moral philosophy for centuries: if you save a life and that life goes on to destroy other lives, what is the weight of your original act of mercy? The surgeon, Dr. Tenma, gives up everything, his career, reputation, safety, to hunt down the monster he saved, and the story forces you to sit with every ethical implication of that choice without ever telling you the right answer. It is not a comic book in the way most Americans understand that term. It is a 162-chapter moral examination disguised as a thriller, and I understand now why my son said he wasn’t ready for it, yet.

Taking Ret to those bookstores and letting him lead me into his world was one of the best things I did this quarter. He’s developing his sense of self in a world of complicated ideas and behaviors, and the fact that he’s doing it through stories about sacrifice and moral courage and the cost of doing the right thing gives me real confidence about the kind of person he’s becoming. I still remember listening to him play with his toys at age 4 and 5. He loved dinosaurs and would always have complicated dramas between a Dad and Son dinosaur and the work they did together defeating evil. His reading material now feels very close to that, and is a delight in a man seeing his son becoming a man.

Orlando — Jeffersonian Dinner

I took my COO, Robert Rodriguez, to a Jeffersonian dinner in Orlando hosted by Pathstone, a firm that we use to manage our money. The topic was the reality of family wealth: how it’s built, how it’s transferred, how it’s lost, and how families talk about it, which in my experience is badly and not often enough.

If you haven’t been to a Jeffersonian dinner, the format is simple: one table, one conversation, no side chatter, no table-hopping, no escape into small talk. You sit and your host asks a series of questions, you listen, you contribute when you have something worth saying, and you hear things you would never hear at a cocktail reception where everyone is performing. The format demands honesty in a way that round-table events do not, because there is nowhere to hide.

Robert is taking the first steps toward being a first-generation wealth creator for his family, and I wanted him in that room for the same reason I take my kids to tables where real conversations happen: context matters. He heard someone describe watching a trust fund dismantle a sibling’s ambition. He heard someone else describe how their family treats capital as a responsibility rather than a prize, and the specific structures they use to reinforce that distinction across generations. He heard a third person talk about the guilt that can accompany inherited wealth, the way it can paralyze you if you don’t have a framework for understanding what the money is for. I could have told Robert all of this in the office. It would not have landed the same way. The credibility of the person speaking matters. The room matters. Investing in someone’s development means putting them in rooms they wouldn’t otherwise be in. That’s a great privilege.

Paris

Emmaline and I spent her spring break in Paris. I’ve been going to Paris for forty years, since I was twelve years old and singing soprano in the Atlanta Boy Choir. I’ve been back more than twenty times. I honeymooned there. I’ve walked every arrondissement, eaten meals I still think about years later, and learned the language after my own fashion—imperfectly, with the confidence that comes from showing up enough times that the imperfection stops mattering.

This was Emmaline’s trip. She is fourteen, old enough to absorb the weight of a city that has been standing for two thousand years and young enough to be genuinely astonished by it. We walked everywhere, because Paris is a city that punishes you if you take the Metro and rewards you if you don’t. I showed her the plaques—the small brass plates bolted to the facades of ordinary apartment buildings that mark where someone was taken by the Gestapo, or where Voltaire died, or where a Resistance cell printed pamphlets. You never see these from the train. You see them on foot, at walking speed, if you know where to look and if you’re willing to slow down enough to deserve them.

She was quiet for a long time after we stood in front of one that commemorated a family of four taken from the third floor of an apartment on the Rue de Rivoli in 1942. The plaque does not say whether they came back. She asked me about it later that night over dinner, and we talked about it for an hour. I have been to Paris enough times that it rarely rearranges me anymore. Watching it rearrange my daughter was worth the entire trip. This was her first real step into the wider world, the first time she encountered history not as a subject in school but as something that happened to real people in a real place that she could reach out and touch. That changes a person. It changed her. She’s been a completely different kid since we got back … no, that isn’t right. She’s been more herself than we’ve seen for several years.

Intellectual Pursuits

Building the Court

This is the section that will either fascinate you or make you think I’ve lost my mind. Possibly both. I’m fine with either outcome.

Coming out of the Optimus mastermind in January, I deployed an open-source platform called OpenClaw and built on top of it what I now call the Court—a set of specialized AI agents, each with a defined role, persistent memory, and bounded authority over a specific domain of my life and business. The naming convention is borrowed from a medieval court, partly because it captures the hierarchy and partly because I think it’s funny.

The Chancellor manages scheduling, operations, and coordination. If you’ve emailed me in the last two months and received a reply about scheduling, the Chancellor may have either drafted it or sent it. The Chronicler handles all written output—newsletters, investment memos, public comments, business correspondence. This quarterly lookback was drafted with his help the. The Compradore manages financial analysis and portfolio tracking. The Constable monitors security and system integrity. The Chamberlain handles household logistics.

These are not chatbots. They are persistent agents with memory that spans sessions, context about my life and preferences that accumulates over time, and charters that define what they can do autonomously and where they must stop and ask me. The Chancellor knows my calendar, my contacts, and the fact that I don’t take meetings on Friday afternoons or from 12-2 during the week. The Chronicler has absorbed my writing voice across hundreds of thousands of words and produces first drafts that require editing, not rewriting. The Compradore can pull portfolio data, build summaries, and flag anomalies I would have missed.

I want to be honest about what this is and what it isn’t. It is not a replacement for thinking. It is a replacement for the mechanical labor that used to sit between a thought and its execution. The gap between “I should write that” and “it’s written” has collapsed from hours to minutes. The quality floor has risen because the agents enforce standards I set once and don’t have to police every day. The system requires real governance, oversight, refinement, and correction. It requires trust. Calibrated, earned, verified trust, but trust nonetheless.

The productivity gain has been the single largest step function in my professional life. I am not exaggerating. This is a system I expect to use and improve for years, and building it was one of the most important investments I made this quarter.

NIST — First Public Comment

In February, the National Institute of Standards and Technology issued a request for information on security considerations for AI agents. I submitted a formal comment about TrustMyBot.ai, the certification infrastructure company I’m building for agentic commerce. It was the first time I have ever responded to a federal call for public comment, and it probably won’t be the last.

The submission focused on what I believe is the critical missing piece in the current discussion: behavioral trust. Mastercard, Google, and Visa have all published frameworks for verifying that an AI agent is who it claims to be and that a human authorized its actions. Good work, all of it. But none of those frameworks address whether the agent is behaving well after it’s been authenticated. An agent can be fully verified, operating within its delegated scope, and still manipulate a counterparty, misrepresent its principal’s requirements, or coordinate with other agents to distort pricing. These are behavioral failures that happen at runtime, after all the identity checks have passed. The current standards don’t account for them. TrustMyBot is built to fill that gap.

The comment period closed March 9. NIST is hosting a public-private convening in April. I plan to be in the room. If you build with agents or deploy them in your business, you should be paying attention to this space, because the regulatory framework is being written right now and the people writing it are listening.

Investments

The Groq exit finalized this quarter. NVIDIA acquired Groq’s inference assets for approximately $20 billion in cash, announced on Christmas Eve. Our checks came back at just shy of a 300% return, with more to come later because of the structure of the deal—NVIDIA set performance milestones tied to the integration of Groq’s LPU technology, and additional payouts are contingent on hitting them. I mentioned Groq in the Q4 lookback when the acquisition was announced; the final per-share numbers settled slightly below the initial markup I reported, but the return is still one of the better venture outcomes in our portfolio history. We wait for the milestones. Patience remains the strategy.

Our broader public market portfolio benefited greatly from me raising cash in Q4, and starting cautiously to redeploy this quarter. We picked up many of the same positions for less money. I took chips off the table in November on the thesis that a meaningful pullback would create a reentry opportunity. Our portfolio was rewarded by that forethought. Patience with dry powder has historically been more profitable than anxiety with a full allocation. I would rather miss some upside than catch a downdraft fully invested. The thesis hasn’t changed.

PulsedIn, a company I first mentioned in Q3 as a stealth investment, is now raising a round that we are proudly supporting. The business uses AI technology to help staff hospitals with nurses, which might sound mundane until you learn that American hospitals are spending billions annually on temporary staffing agencies because they cannot efficiently match their own labor supply to demand. Each nurse fully utilized at a hospital creates between $1-$3m in billable charges of the hospital and the lack of nurses is all too real. PulsedIn is using AI to identify them and empower human recruiters to interact with them in authentic and compelling ways. The average hospital fills those positions in 70 days, PulsedIn does it in 7. I’ve been helping them with business development as well as capital formation, and the conversations with health systems have been encouraging. If you’re in healthcare or know someone who runs a hospital system, I’d love to make the introduction. You’ll be helping everyone involved. And if you’re interested in investing, drop me a note.

Reading

The Book of Elon — Eric Jorgenson

Jorgenson did the same thing here that he did with The Almanack of Naval Ravikant: gathered up years of interviews, tweets, podcast appearances, and essays and curated them into a single readable volume that lets the subject speak in his own words. The editorial hand is light, which is Jorgenson’s great skill and the one most editors lack, he knows when to get out of the way.

The book is organized around Musk’s first-principles thinking, his approach to manufacturing, his tolerance for personal discomfort that would break most people, and his conviction that making life multiplanetary is not aspirational but existentially required. You will either find this clarifying or maddening, depending on where you stand on Musk himself. I found it clarifying. Whatever you think of the man’s politics or his timeline on X, his ability to reason from physics rather than analogy is rare and worth studying.

The chapter that changed me was the one on manufacturing. Musk’s argument is that the factory is the product, that optimizing what you make is less important than optimizing the system that makes it. If the production system itself isn’t being continuously improved, you’re working on the wrong layer of the problem. I read that chapter on a plane back from Scottsdale, two days after the Optimus mastermind, and it crystallized something I’d been feeling for months: the reason I was building the Agentic AI system wasn’t to produce better outputs. It was to build a better production system. The factory is the product. That insight alone was worth the price of the book.

Small Things Like These - Clair Keegan

Small Things Like These is a very thin book that punches above its weight and its punches land with the force of a moral indictment. Claire Keegan has written very good here with strains of Hemingway’s restraint and Paton’s moral awareness. The book doesn’t try to dazzle you with plot twists. Quite the opposite really, it puts a decent man in front of a terrible thing and ask the oldest serious question in the world: what does a good person do when goodness becomes risky?

In Ireland, there were institutions run by nuns called Magdalene laundries. This is where unmarried pregnant girls were sent to give birth and have their children adopted away from them. They earned money by taking in laundry from surrounding towns and institutions to give them work to do and to pay for their keep. By all accounts, the places were terrible in every sense of the word, misery factories that came about because shame was weaponized and tolerated as good works.

Keegan does not attempt to narrate the full historical horror of the Magdalene laundries in some sprawling, panoramic fashion. She does something much more intelligent. She reduces the frame until the you are forced to confront the problem at human size, where all of our real moral decisions are made. The result is that the institutional evil feels more terrible, not less, because it is encountered through cold weather, household worries, church power, small-town silence, and the ordinary calculations by which people talk themselves out of courage.

Bill Furlong is the sort of protagonist literature needs more of. He is not glamorous, not especially powerful, and not fashioned as some grand heroic instrument. He is simply a man with a conscience, which turns out to be a far more demanding thing to wrestle with in literature. Keegan understands that heroism in real life often looks quiet, hesitant, and expensive. That gives the novel its sting. It reminds the reader that most evil does not endure because monsters are everywhere. It endures because ordinary people learn to live beside it and not look at it headlong, or name it as such.

The book’s impact comes from that moral compression. It leaves the reader with the unnerving realization that the great crises of history are often upheld by minor acts of compliance, cowardice, and polite looking away. Keegan never bellows about this point. In fact, she barely raises her voice. That is precisely why it haunts you. The silence around the subject becomes the biggest part of the subject.

As for how good it is, it is excellent in the old-fashioned sense of the word. It is beautifully made, morally serious, and free of bloat. There is not a wasted page in it, nor a misplaced word. Many novels are praised for being “important” when what they really are is loud, long, and dutiful. This one is actually important because it is artistically disciplined enough to make its moral weight felt without ever becoming preachy.

The finest books enlarge the reader’s sense of reality while also sharpening his sense of obligation. Small Things Like These does both. It is brief enough to be read in an evening and grave enough to stay with you much longer than books four times its size. In fact, I’ve thought about it every single day since I read it on a Sunday morning with my coffee. That is not a common achievement. It is one of those rare modern novels that earns the right to be called both beautiful and necessary.

Watching

Firefly (rewatch)

I rewatched Firefly. Again. Because it had been a couple of years and because it remains the best show ever made, and I will take that position to my grave and argue it from the other side if they let me.

One season. Fourteen episodes. No one who has watched it can believe that a television executive would cancel it, but it was a Fox product and they cancel shows like I change socks. Despite all of that, it is still—still—the finest piece of serialized storytelling in the history of the medium. Joss Whedon took the American frontier, collided it with a Chinese/American post-war empire fusion, set the whole thing in space, and populated it with nine characters who are broken and decent and funny in exactly the way that actual human beings are funny when they think no one is listening.

There are so many cliches that could attach to the show and make it insipid and stupid .., but the quality of character development, dialogue and discipline the show brings to this perfect balance of being a small character in a big story is absolute perfection.

Mal Reynolds is one of the best protagonists ever written for television. I know that’s a big claim but I mean it earnestly. He is a man who lost a war, lost his faith, and kept his honor, and the tension between those three things drives every decision he makes for fourteen episodes. The supporting characters are fabulous and I wonder why more shows don’t spend time developing the second rank to make the shows feel more real.

In a word, the show is about freedom, what it costs, what it requires of you, and what you lose when institutions that claim to be helping you start making your choices for you. If that theme sounds relevant in 2026, it’s because Whedon saw it in 2002 and the audience took twenty years to catch up. Watch it. If you’ve already watched it, watch it again. If you don’t like it, we can still be friends, but I will think less of you and I won’t pretend otherwise.

Hoppers

I took Mary-Salter to see Hoppers, the new Pixar movie about a young woman who transfers her consciousness into a robotic beaver to study animals. Mary-Salter loved every frame of it, especially the beaver, and especially the part where the beaver does something brave and the whole theater gasped, including her, at a volume that made the row behind us laugh. The popcorn was bigger than she was. I maintain that ages three to four remain the absolute sweet spot for animated movies: old enough to follow the story, young enough to believe all of it, and young enough that they grab your hand during the scary part without any self-consciousness about doing so. I’ll take it for as long as she’ll give it to me, which probably isn’t much longer, so I’m savoring every frame.

---

Q1 is the quarter of beginnings, and most beginnings don’t survive contact with February. This year they did. Nearly everything I planted is still growing, and some of it has already started producing.

The agentic system runs while I sleep and gets better each week. Robert sat at a table in Orlando where the conversation was about building something that outlasts you, and I watched him take it seriously. Ret handed me a story about a surgeon who has to reckon with the consequences of his own mercy, and the fact that he wanted to share it tells me more about who he’s becoming than any report card could. Emmaline stood in front of a plaque on the Rue de Rivoli and went quiet, and in that silence I watched her step into a world that is bigger and heavier and more beautiful than the one she left. Mary-Salter grabbed my hand during the scary part of a movie about a beaver, and I held on for as long as she’d let me.

This was a quarter for investing in what comes next. The seeds are in the ground. Time to see what grows.