*“Yesterday, I Went to Mars ♡”*

This episode looks at a first hands-on experience with the ROKID AI glasses — a wearable device that overlays information directly onto the lenses.After a few hours of use, a right eye began to hurt. Wearing the ROKID over prescription glasses, the double-glasses situation, may be part of the cause — though whether it's that alone, or something about staring at a display that close to the eye, remains an open question.Beyond the physical discomfort, what stood out most was the UI. A small lag between speaking and receiving a response — probably DeepSeek-based — repeated often enough that the urge to speak to it gradually faded. The touch controls required stopping to think each time, and the fact that familiarity is even required felt like a barrier in itself.There's a comparison drawn to Apple — not to be unfair, given the difference in scale and history, but to name something real. That sense of just being able to use something without thinking is the result of years of small corrections made in response to moments of confusion. With daily-use devices, small friction doesn't stay small. It accumulates quietly until the device ends up on a shelf.A quiet look at what separates hardware that earns a place in everyday life from hardware that doesn't — and a note that, as someone who runs an eyewear shop, the convergence of frames and technology feels like something worth watching closely.

This episode returns to a theme that has come up a few times recently on the blog: companies that seemed finished, only to turn out to be quietly building something in the background of the AI era. This time, the focus is Nokia — a name that still immediately calls up the image of a Finnish mobile phone maker that missed the smartphone wave.The surprise is how completely that picture has been overtaken by reality. Nokia today is an infrastructure company, working in AI networking, optical networks, and wireless access technology for 5G and 6G — with a roughly one-billion-dollar investment and strategic partnership from NVIDIA to go with it.Dell and Intel come into the picture too, as part of the same pattern. Not the companies making the visible, finished products of the AI boom, but the ones supplying the servers, the chips, the manufacturing capacity, and the communications backbone that make any of it run at all.There's a Gold Rush analogy here — the observation that the people who made the most money weren't the ones digging for ore, but the ones selling pickaxes. Watching where value is quietly accumulating in the AI era, it starts to look like the same dynamic.A quiet look at how easy it is to mistake the front of the stage for the whole picture, and what gets missed when the foundation is invisible.

This episode looks at the story behind Dell's recent stock surge — and what that story actually reveals about how reinvention works.It traces how Dell went private in 2013, largely disappearing from public view, then made a $67 billion acquisition of storage giant EMC in 2016 that drew sharp criticism at the time. From the outside, it looked like a company in retreat. From the inside, it was rebuilding its entire foundation.When the AI boom arrived and demand for data center infrastructure exploded, Dell was already positioned at the center of it — not by accident, but because of bets placed quietly, years earlier, away from quarterly scrutiny.There's also a broader reflection on how easily we judge companies, and people, based only on what's visible right now — the "that place is finished" instinct — and how often the real work is happening somewhere below the surface, unannounced.A quiet look at delayed recognition, and the question of whether what matters most isn't immediate results, but the willingness to keep planting seeds in the direction you believe in.

This episode looks at the specific friction that appears before beginning — not a lack of ability, not a lack of motivation, but something closer to the weight of that first instant before anything is set in motion.There's a story here about a piece of work that felt untouchable, and a quiet experiment: just one minute, genuinely meant as only one minute. Thirty minutes later, the work was still going — not because of effort or encouragement, but simply because starting had led to continuing.It touches on the idea that the brain tends to preserve its current state — stillness wants to stay still, movement wants to stay moving — and that the real obstacle is rarely what we think it is.There's also a small reframe drawn from seventeen years of running an eyewear shop: that the times things actually moved forward were almost always the times they started small, and that waiting for conditions to be right has a way of waiting indefinitely.A quiet reminder that motivation often follows movement rather than preceding it — and that shifting a day from "nothing done" to "one minute done" is, more often than not, where everything else begins.

This episode looks at a theory that reframes nightmares not as disturbances to escape, but as something closer to training runs — a neuroscientific idea called Threat Simulation Theory, which proposes that dreaming exists to rehearse dangerous situations before they happen in waking life.The analogy at the center of it: dreams as flight simulators running quietly every night, no real consequences, just the brain running offline tests on fear and survival response.It touches on the evolutionary logic behind this — that for hundreds of thousands of years, individuals whose sleeping brains rehearsed escape and threat response were more likely to survive, and that this pattern left traces visible even now in dream content across cultures. The most common interaction in dreams is aggression, and the dreamer is more often the victim. The threats that appear are overwhelmingly wild animals or unfamiliar men.There's also a recurring nightmare woven through — a car, brakes that won't catch, that straining helpless feeling of pressing and pressing and barely slowing — reread here not as anxiety to suppress, but as the brain rehearsing, night after night, the sensation of an unresolvable threat.A quiet look at what it might mean if the things we most want to erase from sleep are actually the brain being considerate in its own way — making sure that when the moment comes, we're not starting from zero.

This episode looks at a book encountered by chance at a bookstore — *To Everyone Who Couldn't Become a Genius* by Kappy, creator of Left-Handed Ellen — and the idea at its center that stuck: "MP consumption," borrowed from video game logic.The frame is simple but pointed. Use a powerful spell and your MP drops sharply. You win the fight, but you can't move until you've rested. The book applies this to careers — arguing that the core of sustainable work should be built around "high-efficiency cards," things you can do better than most and that don't drain you no matter how long you spend on them.It touches on a parallel idea from Naval Ravikant, and the way the two books seem to converge on the same question: not just "can you do this?" but "when you're doing it, is it eating into your mental HP and MP?"There's also a brief look at how this connects to the AI moment — as more and more things become doable with assistance, the question of what you can *sustain* without wearing yourself down becomes its own separate and more pressing issue.A quiet look at the difference between willpower and efficiency, and why sorting through your own cards — not by what you're capable of, but by what leaves you depleted versus what quietly restores you — might be the more useful way to think about a career.

This episode takes a single scene from the TV series *Billions* — a wife snapping "How are we supposed to live on that?" despite sitting on tens of billions of yen — and uses it as a lens for thinking about why financial anxiety doesn't scale the way we expect it to.It touches on a piece of research suggesting that even people with savings over 100 million yen often still carry money anxiety, and explores why: not because the numbers are wrong, but because the future stays undefined, and an undefined future keeps the mind on alert.There's a close look at what happens when a certain standard of living becomes normal — the house, the travel habits, the schools, the social circle — and why lowering that standard isn't simply a matter of spending less. It starts to feel like rewriting your own sense of who you are.The episode also draws on a description of anxiety as "fear of a vague, possible future threat," using it to reframe the real source of financial unease: not the size of a balance, but the shapelessness of what's ahead.A quiet reflection on what it might actually mean to feel like enough is enough — and why putting that into your own words, rather than chasing a larger number, may be what genuinely helps.

This episode looks at the experience of losing track of time inside a book — that sudden awareness that two hours have passed — and what that feeling might actually be doing for us.It touches on the tension between keeping a wide view of life and the quiet cost of that wideness. Work, family, the direction of a business, old age, the planet — when all of it stays visible at once, it wears on a person.There's a small detail from running an eyeglass shop: during the busy stretches, when new store openings pile up and schedules fall apart, the habit is still thirty minutes of reading before bed. Not planning the next day. Just stepping into a story and letting the view narrow — and finding that by morning, a problem that felt unsolvable looks, for some reason, a little smaller.The episode also gently reframes the word "addiction" — not as something shameful, but as a sign that human beings may simply not be built for constant, wide-open awareness. Someone absorbed in love, or in work, or in a story is narrowing their view in ways that look foolish or unproductive from the outside, but are producing something inside.A quiet reflection on why everyone needs something at hand that can absorb them — and on the idea that seventy percent reality, thirty percent something absorbing, might be just about the right balance for a life.

This episode looks at Intel's sudden 24% single-day stock surge on April 23rd, and what it might actually mean for how AI infrastructure is shifting.It traces the gap between perception and reality — the sense that Intel had become a dinosaur, especially vivid for anyone who switched from an Intel Mac to Apple Silicon and felt the difference immediately.The episode works through a quiet analogy: the CPU as a store manager reading the whole operation, the GPU as a large floor crew built for repetitive, high-volume tasks. For years, AI was dominated by the training phase — the endless-cabbage-shredding work — and that was the GPU's moment entirely.But as AI moves from being built to being put to work, embedded in search and productivity tools and customer systems, what's needed is something that can route and orchestrate and control — and the store manager is suddenly in demand again. Intel's Granite Rapids line, refined quietly over several years, landed at exactly that inflection point.A quiet look at how unglamorous, persistent work can realign with the times in an instant — and a reminder that a picture you were certain of can be flipped by a single earnings report.

This episode picks up a thought from an in-flight newsletter — the idea that, in the end, the people who find their niche are the ones who win — and follows it from 30,000 feet back to ground level.It touches on a pattern visible on YouTube and elsewhere: channels built around something almost absurdly specific ("the most niche harness talk in Japan," "bouldering for beginners in your forties") that end up growing precisely because of their narrowness, while broader efforts to appeal to everyone tend to disappear.There's also a candid look at an early period running an eyewear shop — the phase of wide selection, no particular focus, doors open to everyone — and the quiet realization that came with it: no one could quite say what the store was.The episode draws a line between "choosing a niche" and "turning people away," suggesting they're not the same thing at all — and that a niche, more often than not, isn't something you design in advance but something you notice you've always had, in the topics you research without being asked and the conversations you keep ending up in.A quiet reflection on the difference between trying to reach everyone and being found by exactly the right people.

Claude’s new “Dreaming” feature for Managed Agents feels like an interesting shift in how AI may evolve. Instead of starting from zero every session, these agents can review past interactions during idle time, organize memories, identify recurring mistakes, and gradually adapt to users and teams over time.It’s still an early-stage feature aimed mainly at developers and enterprise systems, but the idea itself is fascinating: AI not just responding, but slowly building long-term context and habits through experience.What stood out to me most is that this changes the feeling of AI from being a simple one-time tool into something closer to a long-term collaborator — something that can gradually learn the tone, preferences, and workflow of the people around it.A few years ago this would have sounded completely like science fiction. Now it’s quietly becoming part of real products and infrastructure.

This episode reflects on a habit that seemed wrong until it suddenly made complete sense — putting ice in beer, something encountered for the first time in Chiang Mai's heat.It touches on the practical logic behind it: daytime temperatures above 35 degrees, a beer that goes lukewarm in under five minutes, and the way ice becomes a form of quality control rather than a shortcut. There's also a small detail about Thai beers like Singha and Chang being brewed slightly stronger to account for dilution — so the crispness holds even as the ice melts.The episode also considers the physical side of drinking in tropical heat, and how slowing down the pace and rehydrating a little as you go made for a noticeably better morning after.What runs through it is a quiet shift in thinking — that the sense of how things "should" be done is often more local than it feels. A practice that might look careless at a bar in Tokyo turns out to be something the people of Chiang Mai refined over a long time, for good reason.A small reflection on the difference between a wrong answer and an answer shaped by a different place.

In this episode, I talk about traveling to Thailand with an Apple Watch Ultra and a Pixel Watch after years of relying almost exclusively on the Garmin Fenix 8.Using different smartwatches side by side made me rethink something I had mostly ignored before: Garmin’s MIL-STD-810 military-grade durability standard.On paper, many modern smartwatches are water resistant and highly capable. But over time, I started noticing the difference between a device that can survive normal use and one that feels truly stress-free in everyday harsh conditions like heat, steam, swimming, saunas, and accidental impacts.I also discuss how small engineering decisions — like Garmin’s button structure — can quietly affect long-term durability in ways that don’t usually appear on spec sheets.It became less about features or performance, and more about the feeling of being able to use something without constantly worrying about damaging it.