*“Yesterday, I Went to Mars ♡”*

This episode reflects on a shop worth visiting — except one staff member's reaction to a simple question about price has made it impossible to walk through the door again.

It touches on that specific moment: asking, a little hesitantly, whether anything came in a lower price range, and the shift in the air that followed. Not a confrontation, not an unkind word exactly — just a tone, and a look, that left a lasting feeling of having wandered into the wrong place.

There's also an honest look inward at Makoto Eyewear, where frames occasionally surprise customers with their prices, and the question of whether that same quiet atmosphere — something leaking through in the manner, even when the words are polite — might be happening there too.

The episode lands on a simple hope: that when someone comes in looking for glasses at a JINS price point and finds something different, the response can be straightforward and unhurried, without any weight behind it. Not because every customer will buy something, but because how they feel when they leave is what stays.

A quiet reminder that pride in what you carry and looking down on people are not the same thing — and that a shop worth returning to has to get both the product and the welcome right.

This episode looks at a Silicon Valley startup called Maxinsights, which just raised 30 million dollars to film ordinary people doing household tasks — loading dishwashers, folding laundry — from helmet and wrist cameras, to generate training data for physical AI and robotics.

It touches on a gap that's easy to overlook: while language models learned from the vast text already on the internet, almost no equivalent data exists for physical tasks. The result is a new kind of gig work in Los Angeles, where two hours of filming yourself doing chores pays eighty dollars.

There's also a reflection on what this means for craft and skilled work — specifically the kind of fingertip knowledge that builds up after handling thousands of eyeglass frames: knowing exactly when to ease off heat on celluloid, or how half a millimeter shift in a nose pad changes everything for a particular customer's face.

It draws a line from that disappearing artisan knowledge — aging masters in Japanese manufacturing whose near-unconscious precision has no one to inherit it — to the possibility that first-person sensor data could preserve and transfer those skills, perhaps not to a human apprentice, but to a humanoid robot.

A quiet look at how the most unremarkable physical gestures of daily life — the ones we perform without thinking — may turn out to be among the most valuable raw materials of the AI era.

This episode reflects on the idea that the ability to "let things pass" may be one of the most undervalued skills for living well — observed through watching children, but arriving quickly at the thought that adults may need it even more.

There's a distinction drawn between endurance and letting things pass, illustrated through an aikido analogy: endurance means absorbing a blow and bearing it silently, while letting things pass means redirecting the force so the damage never lands in the first place.

The episode also traces an unexpected connection between this ability and concentration — the idea that when a word or a moment won't leave your head, your attention has effectively been taken from you, and that focus isn't only about effort but also about what you allow to move through you.

Woven through is the question of what it means to know what matters to you, since without that sense of sorting, everything catches — and with it, the things that don't matter can be redirected without becoming damage.

A quiet reflection on calmness, and how the people who seem unhurried and unworn may simply have, over time, developed a shape that lets the unnecessary pass through.

This episode opens with a familiar feeling: the particular pleasure of browsing specs for a new device, imagining how much smoother everything could run, how many tasks might finally get handled — before the purchase arrives and the floor looks exactly the same the next morning.

From there, it draws a line to a Reuters report on SpaceX raising roughly $40 billion partly for AI infrastructure. What stayed interesting wasn't the scale of the ambition, but what backed it: the quiet, recurring revenue from Starlink subscriptions and routine launches. Ordinary ongoing work, underwriting an extraordinary bet.

It touches on a concept economists call the productivity J-curve — the observation that when electricity first entered factories, simply replacing steam engines with motors changed nothing. Output only improved after entire factory layouts were rebuilt around the new reality. A talk by Sarah Hunter of the Reserve Bank of Australia adds texture here: AI adoption has spread widely, yet around 35 percent of businesses reported no productivity improvement, and roughly 30 percent said only marginal gains. The investment is surging; the results are still catching up.

There's also a practical note on what introducing any new tool actually looks like up close — shelves moved to make room, manuals read, workflows renegotiated with staff, a stretch of feeling busier and more scattered than before the machine arrived.

A quiet look at the gap between imagining a better future and doing today's unremarkable work — and the possibility that those two things aren't as separate as they feel when you're scrolling through a spec sheet.

This episode looks at a moment of quiet realization: that AI's reach may not stop at knowledge work, but could extend into the physical world of human hands.

The trigger is T-Rex, a robotic system announced by Panasonic and others in early November, in which sensors built into all ten fingertips allow a robot to detect force and slippage in real time — adjusting its grip the way a human hand does by feel. Picking up an egg without breaking it, turning pages, rotating a key. A 65 percent success rate across twelve tasks, up from 35 percent with previous methods.

It touches on what that directional shift means more than the numbers themselves. There's a moment of personal reflection on eyeglass fitting — bending a temple, adjusting a nose pad — where the sense of touch has always felt irreducibly human, and what it might mean if even that could change.

There's also a note on the broader structural question: the old logic held that if AI displaced knowledge work, people could move into physical work. T-Rex suggests that division may not hold. Robots don't tire. The obstacles remain enormous, but the boundary is moving.

A quiet look at what stays with you after reading about a technology — not the specs, but the question on the other side of it: if both thinking and touching can be handled by machines, what will people choose to do with their time?

This episode takes a close look at a proposed two-year tax cut in Japan — and the quiet unease that settles in when you've just stepped off a plane from Hong Kong and something about the headline doesn't sit right.

It draws on years of traveling through Italy, Germany, India, Southeast Asia, Macau, and beyond — and the recurring, unsettling realization that Japan, once the expensive destination, has quietly become the affordable one. The image of foreign tourists lining up outside Shinjuku massage parlors at night lands with a particular weight here.

There's also an analogy that cuts through the policy debate cleanly: a patient who needs serious surgery being handed a two-year supply of adhesive bandages. Not cruel, not dramatic — just an honest read on the gap between the scale of the problem and the scale of the response.

The episode doesn't stop at critique. It turns toward what that same money could do instead — funding young people competing on a global stage, building a safety net that makes failure survivable, drawing in talented people from abroad. The contrast is between distributing money and using it as fuel.

A quiet reflection on the difference between defense and direction — and a reminder that when the world is moving fast, the answer probably isn't to make the present a little more comfortable.

This episode looks at CATL's announcement of a battery-swap system for heavy-duty electric trucks — a five-minute full battery exchange, with plans to scale to 10,000 stations — and what that design decision reveals about problem-solving.

It touches on why charging, which works well enough for passenger cars, breaks down completely for freight logistics: heavy loads, long routes, and the reality that hours at a charging station translates directly into money lost.

There's also a comparison to camera batteries and AA cells, and a look at how CATL's answer wasn't to make charging faster, but to remove the waiting entirely by changing the operational structure rather than the technical one.

The episode draws a few other parallels — a vintage eyewear company that stopped competing on new inventory and went all-in on used frames, and an attempt to climb Everest without a mountaineering background by reverse-engineering everything from the summit down.

A quiet reflection on what it means to reach the other side of a wall, and why the opening sometimes comes not from forcing a breakthrough, but from replacing the whole structure of the problem.

This episode looks at the fast-moving smart glasses space, prompted by Warby Parker's announcement of a collaboration with Google and Samsung — a development that arrived just one day after a previous post on the same topic.

It walks through how the category is beginning to split along AI lines: Meta with Ray-Ban and Oakley, Google and Samsung with Warby Parker and Gentle Monster, and Even Realities with their own proprietary system. Each is heading in a slightly different direction, even within the same product category.

There's also a thought that surfaces naturally from all of this — that choosing smart glasses may start to resemble choosing between iPhone and Android, where the decision is as much about which AI ecosystem you want to live in as it is about the frame itself.

One tension the episode sits with is the mismatch in timescales: frames can last for years, but the AI inside them can shift dramatically in a matter of months. A pair that feels like the right choice today might look different six months from now.

A quiet reflection on what it means to wear your AI on your face every day — and why, no matter how capable the technology gets, a frame you actually want to put on still matters.

This episode looks at two recent developments in smart glasses: Subvocal, a device that converts silent mouth movements into text using radio waves, and Ray-Ban Meta Audio, Meta's camera-free version of its smart glasses line.

The Subvocal section walks through how the technology actually works — not thought-reading, but mouth movement detection, with no microphone and no camera. It's an important distinction, and one that still feels a little strange even after the explanation.

The camera-free Meta glasses prompt a more personal reflection: wearing camera-equipped glasses in places like a public bathhouse changing room, and the quiet discomfort of that — not from any intent to film, but from how it might look to the people nearby.

Both technologies are framed as responses to friction that only shows up in real daily use, not on a spec sheet. There's also a note at the end about recently becoming an authorized retailer for Even Realities, and what it means for an eyewear shop to start thinking seriously about smart glasses as a standard part of the business.

A quiet look at how wearable technology edges closer to everyday life — not through bigger features, but by removing the small reasons you'd take them off.

This episode looks at a news story about Google disabling AI features by default for student accounts, and the quiet but interesting detail buried inside that announcement.

It touches on a feature called Guided Learning — where instead of simply giving a child the answer to a math problem, the AI asks questions back and helps them work toward it themselves. A tool built to answer anything, deliberately designed not to.

There's also an honest observation about the pull toward convenience: how even watching a child struggle with homework can trigger the urge to just give them the answer, and how AI makes that shortcut more immediate than ever before.

The episode draws a distinction between what adults and children might reasonably ask of AI — adults looking to hand things off and reclaim time, children still in the process of building the ability to think for themselves. And it notes, with some candor, that the line might matter for adults too.

A quiet reflection on what we might want to keep doing ourselves, even as AI becomes capable of doing nearly everything — and whether a life where AI answers everything is actually the same as a life well-lived.