This week we dive into how my AI operating system experiment has evolved, continue the 2026 Oscar series with Bugonia, and the 2025 Booker series with the prize winner, Flesh. Then, with the benefit of one game’s worth of evidence, full NBA playoff predictions. Let’s dig in.
Technology
Last week I introduced the idea of building a personal AI operating system — a knowledge layer that sits on top of an LLM, fed with everything I consume, so that when I ask it a question I get answers grounded in my material and my framing rather than generic training data. The structure borrows from the scientific method: form hypotheses, run experiments (for now, passively loading articles and transcripts; eventually, actively seeking sources to stress-test specific claims), analyze, update, repeat. I’m running this across five domains: healthcare, building a new billing platform, AI and technology, macroeconomics, and being an AI-native PM.
This week, I started noticing the concept everywhere. Coworkers posting their own versions on LinkedIn. Slack channels filling up with multi-agent setups — the classic “I’ve got a head of engineering agent, a head of design agent, a head of marketing agent” configuration. The idea of an AI operating system is no longer fringe. It’s commoditizing fast. But here’s what I think most of the implementations are missing, and it’s not a novel insight: they’re all single-player games.
What I mean by that is straightforward. Right now, everyone runs AI on their own machine. Whatever they learn, whatever skills they build, whatever insights the system generates — it stays with them. Nothing compounds to the company unless they choose to expose it. Say I create a skill that directs an LLM to build frontend prototypes enforcing strict evaluation criteria. That benefits me, but unless I share it broadly, there’s no company-wide standard for prototyping. You can extend this beyond tasks to knowledge. If my billing team discovers through A/B testing that single-provider mental health practices can handle an exceptions-based UI where AI generates and submits claims in the background, that insight could reshape how the clinical team builds the notes layer upstream. Traditionally, that knowledge travels through a meeting or a document — and only reaches the people in the room. But imagine if it became shared context that any team’s AI system could draw on. Now multiply that across every team running experiments in their own domain, all contributing to a shared knowledge base where LLMs cross-pollinate insights across the company. That’s the multiplayer game. That’s how the flywheel kicks in.
I believe this is where AI in the workplace is heading. Not individual contributors with personal copilots, but company-wide operating systems that compound insights across every team and domain. The single-player game is the training phase. The multiplayer game is the endgame. This is also how companies build defensible IP in an AI economy — not through proprietary code, which depreciates faster than ever, but through compounded institutional knowledge that no competitor can replicate.
But here’s the tension I’ve been sitting with: if the company benefits from shared knowledge, what does the individual protect?
I think the answer is that there are two layers, and they need a clear boundary. The first is the professional layer — the skills, systems, and insights you build for your company. This is the multiplayer contribution. You codify workflows, share what works, feed the company’s collective intelligence. The second is the personal layer — your OS, your knowledge base, the lens through which you see your domains. This is what travels with you from company to company. It’s not a file someone can copy. It’s a perspective built from years of reading, connecting, testing, and refining your own hypotheses across the topics you care about.
These two layers interact constantly. Your personal OS makes you better at work. Your work generates inputs that feed your personal OS. But the boundary matters. The professional layer produces outputs — skills, systems, insights — that belong to the company. The personal layer produces the intelligence that generated those outputs, and that belongs to you. When you leave for a new job, you don’t take the company’s skills files. You take your OS — your sharpened understanding of healthcare, technology, macroeconomics, whatever your domains are — and you plug it into a new context. You’re immediately more valuable because you arrive with compounded knowledge, not a blank slate. If you’ve spent three years sharpening hypotheses about where value concentrates in healthcare, which regulatory trends matter, and which business models survive disruption, you don’t need three months to ramp up at a new healthcare company. You walk in with a point of view.
This is where I think the real differentiation lives for knowledge workers. Not in hiding what you know from your employer — that’s a losing game and a paranoid one. But in building a personal intelligence layer that compounds independently of where you work. The company gets your best work. You keep the engine that produces it.
I also want to name something I’ve been thinking about more broadly, because I think it’s a belief that will shape how I approach AI going forward: the purpose of this technology is not to simplify your life to the point where you do less. It’s to enrich it to the point where you can do more of what you’re passionate about. The knowledge that used to require expensive coursework or a personal tutor to acquire, the articles you couldn’t fit into your day, the connections between ideas you didn’t have time or the ability to draw — AI makes all of that accessible. It doesn’t replace the curiosity. It feeds it. I’m in a phase of deep curiosity right now, which hasn’t always been the case. There was a long stretch of my twenties where I barely read a book, and now reading is part of the texture of my week. I’m grateful for that spark, and I want to build systems that protect and compound it rather than automate it away. And that enrichment is what produces the thing that actually matters: richer mental models, which lead to better judgment, which is the scarce resource in a world where execution is cheap.
Movie
This week I watched Bugonia, a 2025 black comedy thriller directed by Yorgos Lanthimos, starring Emma Stone and Jesse Plemons. I wasn’t deeply familiar with Lanthimos’s work going in, but I gather he’s known for the absurd — and Bugonia delivers, though it saves its most committed absurdism for the very end.
The film centers on Teddy (Plemons), a conspiracy-obsessed beekeeper who, with his cousin, kidnaps a powerful pharmaceutical CEO (Stone), convinced she is an alien empress plotting Earth’s destruction. Teddy subjects her to paranoid interrogations in a basement while the CEO deploys her corporate cunning to manipulate the duo, leading to a series of tragic accidents. The film culminates in a revelation: the woman is a genuine extraterrestrial who, after witnessing Teddy’s violent delusion and humanity’s broader cruelty, decides the species is a failed experiment and triggers a global extinction event.
Plemons is excellent here. He shifts from his trademark silent menace to something far more vocal, erratic, and physically exposed. You rarely root for the characters he plays, but as Teddy Gatz he lands on something like a lovable loser — tragic and pathetic in equal measure. The thematic engine is the conspiratorial echo chamber, which the film names explicitly. Someone so deep in their own theory that contradictory evidence only strengthens their conviction, because the rest of the world is brainwashed. It’s tragic and a little frightening. We can laugh at it, but the story knows it’s telling you something sad.
There’s also a deeper thread on the human condition. When Stone’s character reveals herself as the alien, she reflects that she promised herself she wouldn’t end up like this — initially read as being corrupted by corporate greed, but ultimately meaning corrupted by the human condition itself. The film’s verdict is bleak: humans spent their existence at war with each other, let greed destroy the planet that houses them, and are, in the alien’s framing, a flawed experiment defined by violence and narcissism. The final image — bees returning to their hives after humanity’s extinction — suggests that regeneration only begins once the carcass of human civilization is removed.
I respect the thematic ambition, and Plemons and Stone are both doing real work — Stone went fully bald for this role, which is the kind of commitment that elevates everything around her. But the film resolves too completely. The twist announces its meaning rather than letting it settle. Bugonia answers everything — the alien was real, humans are irredeemable, the bees inherit the earth — and once it answers, there’s nothing left to carry with you. The themes are well-understood, especially now, in a world already saturated with climate grief and algorithmic echo chambers. The film confirms what you already feel rather than making you feel it differently. The best films in this genre trust the audience to sit with discomfort. Bugonia doesn’t — it explains itself, and in doing so, lets you off the hook. Strong craft, strong performances, but ultimately a film I admire more than one I’d return to. 3/5.
Books
This week I continued the 2025 Booker list and went straight to the top — the prize winner, Flesh by David Szalay. Full disclosure: it was also the first book Libby let me borrow without a months-long waitlist, so the selection was equal parts ambition and availability.
Flesh follows the life of István, a shy Hungarian man, from an awkward adolescence on a humble estate to an improbable ascent into London’s super-rich elite. Written in a detached, minimalist style — monosyllabic dialogue, clipped exchanges, the classic Eastern European masculine register where strength is measured in what you don’t say — the novel tracks István across decades and geopolitical upheavals as he drifts through roles: soldier, bouncer, driver for billionaires, and eventually a transactional marriage that brings him unimaginable wealth.
I have to say, I loved the writing. Sparse yet vivid. István answers questions with “yeah” or “okay” or, if you’re lucky, “I know.” There’s an element of authentic conversation in this — you’ve met this man — but underneath it is something more painful: the sense that István has never grown past the shy, closed-off kid he was at the beginning. He attracts women, always older, usually unhappy in their marriages, and these women shape his life more than anything he does with agency.
The author said in an interview that his goal was to show István at the end as someone who resembles who he was at the beginning — but with growth you can’t quite put your finger on. I find that so profound, and so true. This is what living is. You carry the full weight of your experiences, and that weight changes you in ways that resist easy articulation. You’re recognizable, but different. The texture shifts even if the shape holds. Readers get to take the broad, third-person view of a man’s entire life — flawed and unspectacular — and try to understand him. In that way, the novel succeeds on its own terms.
There is one passage that hit me deeply. István’s young son tells him he wants to be a fireman when he grows up, and István smiles. Then he has a quiet internal moment where he realizes that his son will never actually be a fireman — that the wealth he’s accumulated has opened a world of choices his son doesn’t even know exist yet. Unlike István, who had nothing and no options, his child will have limitless ones.
I felt this personally, as a son of immigrants. The dream of generational wealth isn’t always about the dollar amount you pass down. Sometimes it’s knowing that your children will never have to work the jobs you worked, that the ceiling you hit becomes the floor they start from. I can imagine what it means for immigrant parents to look at their kids and know they can dream bigger. In István’s case it was money that opened those doors. But for immigrant parents who don’t pass down extreme wealth, the happiness comes from knowing that their hard work earned enough stability to give their children the one thing they never had — options. And as the child, you often don’t realize what was given until much later — the weight of that sacrifice only becomes visible in hindsight, when you look back and understand that your starting line was someone else’s finish line. That passage broke through the novel’s carefully maintained detachment and became something raw.
The novel doesn’t let István keep any of it. His wife and son die in a car accident. His stepson Thomas — who never accepted him, who saw through the transactional marriage from the start — inherits everything once he comes of age and sues István for what he took. The book ends with István bankrupt, back in his village, his mother recently passed, living out his remaining days alone. The same village he left as a young man with everything in front of him.
And yet — recognizable but changed. He started with nothing, accumulated everything, lost everything, and returned to where he began. But the man sitting in that village isn’t the boy who left it. The weight of what he lived through — the luxury, the loss, the one moment where he chose to save Thomas’s life when letting him die would have secured his fortune — that weight is the novel’s answer to whether a life had meaning. That choice — saving the person who will take everything from you — is the growth Szalay was talking about. The teenager who opened the novel couldn’t have made it. István can’t articulate why he called the paramedics. He probably doesn’t understand it himself. But the reader does, and that gap between what István knows about himself and what we know about him is where the novel lives. You may not be able to name the growth. But it’s there, in the texture, and it’s what separates having lived from merely having existed.
4.5/5.
NBA Playoffs
As of this writing, the first game of every series outside of Pistons-Magic and Spurs-Blazers has been played. So I have one game’s worth of evidence — enough to be dangerous, not enough to be right. Given my play-in predictions were littered with misses, don’t mistake confidence for accuracy here.
Western Conference
Thunder vs. Suns (1 vs. 8). Game 1 showcased that ferocious Thunder defense. I don’t see how the Suns find enough offense to compete, let alone win a game. Thunder in 4.
Lakers vs. Rockets (4 vs. 5). An interesting matchup at full strength, but both sides are diminished — Lakers without Doncic or Reaves for potentially the entire series, Rockets without Durant though he’s expected back at some point. Game 1 showed Houston looking flat, and reports all season point to chemistry issues and a disconnect between the players and coaching staff. There’s an air of let this season be over already. Even so, the Lakers are running on a 41-year-old LeBron as their best available player. This grinds to a long series. Lakers in 7.
Nuggets vs. Wolves (3 vs. 6). The marquee first-round matchup — easily a second-round series, arguably a Western Conference Finals. Game 1 showed enough resolve from Denver to make me confident. Edwards still needs to be more efficient in the biggest moments, but this is ultimately a tough matchup for Minnesota. The Wolves’ streak of conference finals appearances ends at two. Nuggets in 6.
Spurs vs. Blazers (2 vs. 7). The Spurs’ turnaround from top lottery team to second seed has been the story of the season. This is their coming-out party — the Warriors-Clippers moment that announces a team has arrived. Wemby and company are too much here. Spurs in 5.
Thunder vs. Lakers (second round). I expect Doncic and Reaves to be back by now, which makes this tougher for OKC than it looks on paper. At full strength, the Lakers were hitting their stride, and playoff Luka is real. But the Thunder have the personnel to defend them, and the Lakers’ defensive limitations with their current core are well-documented. Thunder in 6.
Spurs vs. Nuggets (second round). Another marquee matchup. Denver’s road to a title runs through a gauntlet. Wemby and Jokic already showed us what this rivalry looks like a few weeks ago, and it was spectacular. This is the series where the league learns the Spurs are here to stay — the beginning of something special. But it’s not yet their year. They need more three-point shooting, more playoff reps. They’ll learn from this one. Nuggets in 7.
Thunder vs. Nuggets (WCF). Rematch of last year’s incredible series — the best of the playoffs, save for a deflating Game 7 where Jokic was gassed and Denver didn’t have the depth. The final game will be more competitive this time, but the Thunder are simply too deep. Thunder in 7.
Eastern Conference
Pistons vs. Magic (1 vs. 8). Cade and Duren have made huge strides, but Detroit didn’t address three-point shooting at the deadline. They won’t need it against a Magic team that’s been wildly inconsistent. Pistons in 6.
Cavs vs. Raptors (4 vs. 5). Trading your younger All-Star guard for the older James Harden looks questionable on paper, but Garland was becoming injury-prone and the defensive fit with Mitchell in the playoffs was never right. I liked the move. It won’t be tested here — Toronto doesn’t have enough. Cavs in 5.
Knicks vs. Hawks (3 vs. 6). Atlanta has moved on from the Trae Young era and has strong lottery odds coming, but they’re building for the future, not competing in the present. Not enough gas against New York. Knicks in 5.
Celtics vs. 76ers (2 vs. 7). I’ve seen this movie before. The prime Embiid version. The James Harden edition. Now the Paul George, Edgecombe, and Maxey era. Philadelphia simply cannot beat Boston in the postseason. Celtics in 5.
Pistons vs. Cavs (second round). The second round is where the East gets good. Division rivals, and I expect a physical, volatile series. I may have too much faith in Cleveland here, but their depth wins out. Cavs in 7.
Celtics vs. Knicks (second round). The East Coast rivalry. It wasn’t long ago that the Knicks were the preseason Eastern favorite — they won the NBA Cup, they were riding high, and even after setbacks the talent alone kept them in the conversation. Then Tatum came back from the Achilles ahead of schedule and has looked sharp. Combine that with strong seasons from Brown, White, and Pritchard, an overperforming supporting cast, and a stellar coaching job from Joe Mazzula, and the Knicks are facing a different animal. Boston gets revenge for last year. Celtics in 7.
Celtics vs. Cavs (ECF). We’ve seen this matchup before. Mitchell alone couldn’t get Cleveland past Boston. Is playoff Harden the difference? No. Celtics in 6.
NBA Finals: Thunder vs. Celtics
This is the Finals matchup we were robbed of last year. The Pacers were a worthy replacement — they might have won it if not for the Haliburton injury — but this is the one the league wanted. Both teams are built from the same blueprint: great front office, great coach, MVP-caliber superstar, deep supporting cast, defensive identity, well-run system. Everything I look for when picking a champion. The Thunder just have a shade more, and they have the MVP. Thunder in 6.