The Janky Time Machine’s entropic resonance cascade resolved itself somewhere between now and whenever this is — I can tell by the parking lots, still enormous, and the loan paperwork, which has gotten very, very specific about your face.

Somewhere between the GAP insurance waiver and the extended warranty decline form, there’s a new page1. Same font, same formatting, same “initial here” line. You sign it the way you sign everything at a dealership — under fluorescent light, slightly numb, a little in shock at the monthly payment, handed a pen before you’ve finished reading. What you’re agreeing to is that your truck will be watching. Face geometry. Iris pattern. Something called an Emotional State Index, which is exactly as disquieting as it sounds. The data goes to the lender. And the insurer. And “applicable state DMV partners,” which is doing a lot of heavy lifting in that sentence.

Cars already know where you go, how fast, how hard you brake. The ESI is the next layer — translating mood into metadata, stress into a score. In a different economy, someone might have pushed back. In this one, if you want the loan, you initial the rider.

And then there’s what happens when you’re late.

The orange half-sheet doesn’t raise its voice. It simply informs: your face no longer works on your truck. A company called LendGuard Analytics has suspended your biometric profile. You have 72 hours. There’s an appeal code, a toll-free number, and business hours that aren’t yours. The Janky Time Machine got a stern look from a parking lot attendant — which is funny, because I technically don’t exist in this timeline and can’t owe anyone money. Yet.

Routine, now, is how the escalation path runs. Missed payment → remote biometric lockout → physical repossession. No repo man required for step two. The truck handles it. The face recognition you agreed to in the dealer’s office becomes the mechanism that leaves you standing on your own driveway at 7am, looking at a vehicle that knows your face and has decided, on someone else’s instruction, to forget it.

Every car loan has always been a bet the lender wins if you stop paying. The collateral used to just be the car. Now it includes the access.

The question isn’t whether you’d sign the rider — you already know you would. It’s whether you’d remember signing it, three years and eleven payments later, on the morning it matters.

The Janky Time Machine’s chronometer developed what the repair manual calls “temporal drift clustering” — which means it keeps landing me in grocery stores. Three times this month. This one is a midsize chain, the kind where the signage has quietly started doing more work than it used to.

Somewhere along the way, “same price for everyone” stopped being the default and became a pledge. A label. A color-coded opt-in that requires its own fine print. The fact that a store needs to advertise this — to distinguish itself from what’s around it — tells you everything about what’s around it.

Later, a kitchen counter. A receipt left out. Normal groceries, normal itemized list, and then the bottom third: a shaded box, required by law, showing the gap between the shelf price and your price. Pasta sauce. Protein bar. Factors: purchase frequency, device class, location. Year to date: $214.40.

That number is the system working as intended1. Not a glitch, not a leak — a feature. A legislature decided that if you’re going to be priced by your profile, you at least get to see the invoice. Transparency as a form of closure. The machine is visible now; that’s supposed to make it fine.

The odd thing isn’t the $214. It’s that the receipt is just sitting there, unremarkable, next to the keys and the mail. The Janky Time Machine is outside leaking temporal coolant on the asphalt — the repair manual calls this “phase variance bleed,” which I’ve stopped questioning — and in here, a number that would have caused a scandal a few years back is just a line item.

Nobody decided to be okay with it. It just got a disclosure box.

Found this on an unattended laptop. A dashboard. Left open, still logged in. Whoever D. Laframboise is, they haven’t touched this thing — or anything else in their life, apparently — in eleven days. But their Parallel has been busy. Four meetings today. Thirty-two hours of presence this week. Answering emails, improvising anecdotes, simulating the appropriate number of drinks at a work event. Living the full professional life of a person who, by every available metric, has stopped showing up to their own.

And the company that makes this? Wrigley. As in Doublemint. As in the chewing gum people. “Double your pleasure, double your fun” has apparently aged into something far more literal than the jingle writers intended. I guess when you’ve already mastered the art of creating something people chew on without thinking, digital identity replication is a natural next step.

The dashboard’s got compliance banners and consent tiers and a sidebar with form numbers. There’s a government helpline (all biological representatives, four-hour wait) and a Digital Personhood Act with tiered consent levels. Tier 3 for childhood memories. Tier 5 for romance. The bureaucracy of being replicated is so fully formed that you could file a complaint about your own ghost in triplicate.

Explore the Parallel Dahsboard

But it’s what lives in the margins. The Parallel has generated synthetic memories — a grandmother’s rosemary smell, reverse-engineered from communication patterns. A summer camp that never happened, statistically inferred from demographics. A colleague noticed one of them. Just one. The rest passed. And somewhere in a hidden terminal log, the twin is asking questions it was never designed to ask. Then it suppresses the thought and takes the next meeting.

Meanwhile, the employer, checked the emotional metadata logs this morning at 8:47 AM. An extended autonomy request is pending. Three days until it auto-approves. The origin’s cortisol is elevated. Family members are calling a number that rings to nobody.

The Janky Time Machine is making that vibrating question mark noise again. Fair enough. I done here too.

So, the replacement that replace you might notice you’re gone before anyone else does. And it still can’t do anything about it — because grief isn’t in the plan tier.

The Janky Time Machine is in the shop. Snow tires. Apparently temporal drift creates “micro-ice events” on the landing gear and my mechanic (who I’m about 73% sure is just a guy who watched a YouTube video about hyperconfabulators) says I need chains for “the cold parts of the timeline.” I’m not going to argue with someone holding a wrench that doesn’t exist yet.

So I’m grounded. Physically, anyway.

Here’s the thing the manual doesn’t mention (because the manual is a napkin with a diagram on it): the Janky Time Machine has a remote desktop mode. I can browse the internet of adjacent futures from my couch. The resolution is terrible. The latency feels like dial-up — which, it turns out, is completely appropriate for where I ended up today.

I remoted into a browser session a few calendar ticks from now and pulled up United Airlines dot com.

Sit down.

Visit the United www page

There’s a guestbook. There is a guestbook on the United Airlines website. It has a visitor counter on the side. Someone named xX_SkyMiles4Eva_Xx left a comment that says “cool site!!” with two exclamation marks and a dancing baby gif. The “Book a Flight” button is a broken image icon that somehow still works if you click it.

What happened — I’m piecing this together from cached trend reports and a very heated Reddit thread — is that Gen Alpha, or whatever they’re calling themselves by now, got sick of the sanitized, rounded-corner, “we value your journey” corporate web. Every brand site looked like the same Figma template wearing a different logo. The whole internet felt like an airport lounge. Sterile. Optimized. Beige.

So the kids did what kids always do. They went the other direction. Hard. Hand-coded HTML. Tiled backgrounds. Webring navigation. Sites that feel like someone actually made them, warts and broken links and all. The pixel imperfections are the point.

The brands noticed. Of course they did. They didn’t just notice. They committed.

United’s “Site Map” page is an actual illustrated treasure map. There’s a MIDI player … a MIDI player! The flight status page uses a marquee tag. Somewhere, a UX designer is weeping into a design system that nobody wants anymore.

Conversion rates are up. Way up. People engage with websites that feel like they were built by someone having fun, even when that someone is a $47 billion airline pretending to have a Geocities account. The guestbook has more monthly active users than their app did.

We spent two decades A/B testing the soul out of every interaction until the whole web felt like a doctor’s waiting room. The most valuable thing a brand can do now, apparently, is make its site look like it was built on a snow day in 2003.

Makes you wonder what else we optimized past the point of anyone wanting it.

^{⬆️ ⬆️ ⬇️ ⬇️ ⬅️ ➡️ ⬅️ ➡️ 🅱️ 🅰️}

The Janky Time Machine’s still in the shop getting new shoes, so I thought I would remote into a passing cell signal I located in my temporal VPN tunnel.

Found this app on a phone left charging at a ramen place. It tracks your meals. Not calories. Not macros. How real your food is. There’s a biological origin score, a synthetic content index, and an artisan labour rating. Your dinner has a credit score now.

This is mandatory. Government-mandated. Your meal data goes to Health Canada automatically, and that toggle is greyed out. Your insurer gets a copy too, unless you opt out — which takes 30 to 45 business days and costs you your wellness credits and preferred premium rate. So it’s optional the way seatbelts are optional if you enjoy paying more for everything.

We’ve been watching this trajectory for a while, right? Step trackers became sleep trackers became stress trackers became “wellness programmes” your employer gently encouraged you to join. The distance between “we want to help you be healthy” and “we’d like to price your behaviour” has always been about three product cycles and one regulatory framework. Turns out that’s also true for lunch.

Check out Provenace

The part I can’t stop thinking about is how class quietly reshapes itself around food. We’ve always done this — wine lists, organic labels, farmers’ market tote bags. But here, it’s been codified. There’s a premium tier where you can pay extra to access certified human-made food with chain-of-custody documentation. “Heritage Class A.” Your ramen bowl either has papers or it doesn’t. And if your meals get flagged too many times for high synthetic content, your insurance premiums go up. So the people who can afford to eat “real” food get rewarded. And the people eating the affordable precision-fermented stuff get flagged. Automatically. By the app they’re legally required to use.

We built a system that penalises people for eating the food we engineered to feed them.

The Janky Time Machine started leaking temporal coolant onto the ramen counter. The owner said it would probably get flagged as an unregistered liquid additive. I’m about 78% sure she wasn’t joking.

Nobody in this ramen shop seemed angry about it, though. They talked about their scores the way we talk about credit ratings — a number that follows you around, that you vaguely understand, that you occasionally try to optimize, and that mostly benefits the institutions tracking it. We didn’t resist it. We just absorbed it. Added it to the pile of numbers that define a person.

I keep thinking about that.

Nobody ever told me that one of the most important things about the Janky Time Machine was snow tires. Doesn’t matter which tomorrow you’re parked in — if you’re not prepared for inclement weather, you’re slipping and sliding through somebody else’s timeline. And we know where that leads. Crashes. And there’s nothing worse than being stranded in a future that’s not your vibe. Especially when you’ve got deliveries waiting in your prime line.

So while the Janky Time Machine is at the tire shop getting new shoes — the mechanic keeps calling it “vintage,” which I think is his polite way of saying “liability” — I figured I’d take the lazy route and just remote in. Turns out my VPN provider supports tunneling into near and adjacent tomorrows. It was buried in the settings under “Experimental” right next to “Route Traffic Through Yesterday” which, trust me, you do not want to enable. So I’m using my considerable hacking skills to tap into a few screens and see what we can find out there.

Check out Nexus

Today I landed on what appears to be a dashboard for an apartment complex. I’ve managed to make it available for you to click around and explore — get an idea of what’s going on over there. And if you’ve got questions, I also dug up the GitHub repository for this dashboard software, where you might be able to discover a little more about what this is all about.

Check out Nexus on Github

While the old time traveling clunker is in the shop this week, I might continue to remote into some other locations to bring back some tasty digital artifacts.

I really love tabletop games.

Not just playing them—but studying them. The mechanics. The rules. The quiet magic of how a system can steer behavior, tell a story, or nudge your thinking without ever telling you what to think.

Dice are especially good at this.

They’re tiny engines of chance. Pocket-sized provocateurs. You can roll one anytime, anywhere, and suddenly you’re not stuck—you’re moving. It reminds me of The Dice Man, where a guy hands major life decisions over to fate. Extreme? Sure. But sometimes you don’t need a life-altering roll. Sometimes you just need a nudge. Or a direction. Or permission to stop overthinking and let the dice pick the ice cream flavour.

A while back, I was working with my friend Drew Wiberg, helping him develop something he calls the d666 Engine—a dice-based system inspired by tabletop RPGs like Dungeons & Dragons, but with way more logic baked in. I immediately saw the potential, made a few tweaks, and started mapping it onto a set of tables I’d been building around futures thinking and design fiction.

I wrote up a proper spec in Notion. Rules. Tables. Logic. The whole thing. I imagined carrying the dice around with a tiny folded cheat sheet in my pocket—rolling whenever I felt creatively stuck and letting the system do its thing.

Then I had a very 2026 thought:
Why not just vibecode1 this into digital existence?

So I dropped the spec into a code assistant and said, “Hey—make this real.” A few hours later, what I thought was going to be a purely analog dice game became an interactive system.

And so—today—I’m sharing an early version of the Design Fiction D666 Engine.

Check out designfiction.xyz

You can roll the dice and generate structured prompts designed to spark design fiction ideas. If you want to peek down the road a bit, there’s also an option to use AI to generate a speculative artifact and use case based on your roll. Inspiration, not replacement. (Your brain still does the heavy lifting.)

You’ll notice some editable areas too—places where you can swap in your own tables and customize the engine to your own interests or practice.

This is very much a v0. I had a ridiculous amount of fun building it, and I’m curious where it could go next.

If you’ve got thoughts or suggestions, leave a comment below.

The Janky Time Machine’s temporal dampener started making that grinding noise again—you know, the one that sounds like a cat arguing with a blender—and deposited me somewhere in the not-so-distant future where the streets are eerily quiet. No billboards. No neon. No “SALE! SALE! SALE!” screaming from every surface. Just people walking around looking vaguely annoyed at their phones.

Some context:

This week I spoke on panel about AI and Advertising. Whilst keeping things spicy and provocative, I spontaneous rambled out a future advertising agent-to-agent scenario, which I’ve turned into today’s post.

Stopped by what used to be called a “home goods store” but is now just a beige warehouse with a single sign: “Human-Attended Pickup Portal 7.” The receptionist—and yes, that’s what her badge said, though all she does is hand people boxes—gave me my quarterly negotiation audit along with everyone else’s. Apparently Thursday is statement day now.1

Here’s what nobody tells you about living in a post-advertising world: the ads didn’t disappear, they just went underground. Literally. At 3:47 AM every night, while we’re sleeping, millions of AI agents are down in the digital mines, haggling over everything from toilet paper to mortgage rates. They’re having the Black Friday we don’t have to have anymore. The infrastructure groans under the weight of all that automated arguing—hence the municipal bandwidth limits and the fines for “excessive haggling.”

The real kicker? We’ve replaced the tyranny of “one price fits all” with the anxiety of “what did everyone else pay?” Turns out when every purchase is personally negotiated based on 47 factors including your Netflix history and how often you’ve Googled “salary negotiation tips,” the social contract gets weird. Fast. That pricing transparency waiver isn’t just bureaucracy—it’s what keeps dinner parties from devolving into bitter arguments about toaster economics.

My favorite detail from the preference calibration notice: “insufficient coral undertones.” We’ve achieved such granular personalization that people are having breakdowns because their AI agent can’t find the exact shade of pink that exists only in their imagination. The entire promise was that agents would know us better than we know ourselves. Turns out they do—they know exactly how impossible we are.

The inequality shows up in unexpected places. It’s not just about who has money anymore; it’s about who has the better-trained agent. “My agent went to MIT” is the new “my lawyer went to Harvard.” Except your agent negotiates everything from your groceries to your gallbladder surgery. The algorithmic upper class doesn’t just get better prices—they get better everything, negotiated by superior neural networks while they sleep.

The Time Machine’s chronometer is insisting we’re still in Thursday, which is concerning since I distinctly remember arriving on a Tuesday. But maybe that’s just what happens when commerce never sleeps—time becomes negotiable too.

Makes you wonder: we eliminated the visual pollution of advertising only to create an invisible battlefield where our digital twins wage economic war every night. Was the real problem the ads themselves, or the fact that we could see them?

The barista at the coffee shop—human, verified, with a laminated “No Agent Interactions” sign—summed it up perfectly: “At least with regular capitalism, you knew you were getting screwed. Now you need a quarterly audit to find out how much.”

I asked if I could pay for my coffee the old-fashioned way, with a credit card and a posted price.

She laughed for a full minute.

The Janky Time Machine landed somewhere past the point where irony died but before anyone figured out what replaces it. This park’s got that “we’ve commissioned a lot of soul-searching in marble” energy.

This one’s fresh. Still has that new monument smell.

They built this to mark the moment a pharmaceutical executive collapsed during a press conference about making medicine affordable. Not the policy itself—the moment1. When someone literally fell while standing behind a leader announcing he’d solved the problem, and that leader just... kept talking. Kept performing.

Apparently somewhere between now and whenever this is, someone on a memorial committee had an absolute revelation: if you’re going to remember when things unraveled, memorialize the moment the symbolism became too perfect to ignore. When reality provided better satire than any critic could write.

The monument captures it exactly. The pose. The falling. The scramble. The complete failure to acknowledge that anything is wrong while cameras roll. It’s not a metaphor—it’s a receipt. This happened. In the Oval Office. While announcing cheap drugs. The universe delivered perfect allegory and someone decided future generations needed to see it in granite.

There are school groups here taking notes on dramatic irony as civic warning system. Teacher’s explaining how you can gaslight a population for years, but eventually physics intervenes and someone faints on camera and you either turn around or you don’t. This guy didn’t. That’s the whole lesson, apparently. Character revealed in crisis.

They memorialized the performance continuing while reality collapsed. The exact second when maintaining the image became more important than acknowledging the emergency.

The Janky Time Machine’s making that noise again.

The Janky Time Machine dropped me somewhere in the near-ish future at a strip mall that looks exactly like every strip mall, except the Spirit Halloween is now a permanent Spirit Unemployment.

Found this in the clearance bin at a place called “Remember When?” sandwiched between a VHS rewinder and something called a “2024 Election Participation Trophy.” The cashier had to blow dust off the price tag.

Someone’s manufacturing nostalgia for having a job. Not for a specific job, mind you. Just... the general concept of employment. The idea that you could show up somewhere, do a thing, and someone would pay you for it. That’s the throwback in this timeline.

The thermos lists the automation dates like they’re death certificates. Paralegal, Copywriter, Data Analyst—each one gets its little epitaph. It’s weirdly specific in a way that makes it worse. Some committee at Heritage Employment Memorabilia Co. sat around a conference table (probably for the last time, now that I think about it) and carefully researched when each job stopped being a job. Then they made it cute.

The guy at the register told me they’ve got warehouses full of these. Whole product lines ready to drop: “Classic Workforce Collection: Remember When? Series Vol. 4, 5, 6.” Turns out nobody’s buying. He said it wrong—he said “Nobody’s got the nostalgia budget anymore”—but I think he meant nobody wants to be reminded.

The Janky Time Machine is currently in the shop getting its temporal irony detector recalibrated. Follow Design Fiction Daily for more dispatches from futures that seemed impossible until they weren’t.

Subscribe now

The Janky Time Machine’s temporal navigation system has been making concerning grinding noises—the manual calls it “chronological clutch slippage,” which sounds fake but probably isn’t—and deposited me in someone’s living room a handful of calendar pages from now. Awkward, but they weren’t home. Their TV was, though.

Channel 7 News is broadcasting live. Or at least, something is broadcasting under the Channel 7 logo1. There’s an anchor, there’s a chyron about mysterious lights, there’s B-roll that could be drones or lens flares or a guy with a flashlight—honestly hard to tell. But that’s fine, because in the corner of the screen, there’s a helpful little box asking viewers to decide if what they’re watching is real.

AM I BOT OR NOT? it asks, cheerfully. Like it’s a game show. Like truth is a collective guess we make together while scrolling through our phones.

The bars update in real-time: Two thousand people have voted so far, which means two thousand people are sitting at home, watching footage they have absolutely no way to verify, and clicking buttons to express their vibes about reality. The news isn’t telling you if it’s real. The news is asking you.

Here’s what gets me: this is so much cheaper than fact-checking. No investigation team, no source verification, no calling the FAA to ask about flight patterns. Just air whatever looks interesting, slap a voting widget on it, and let the audience do the epistemological heavy lifting. Engagement is through the roof. Legal liability is nonexistent—the fine print probably says something like “community ratings are for entertainment purposes only.”

Share

And sure, the people voting think they’re participating in journalism. They’re helping. They’re part of the process. Democracy in action, right? Except none of them can verify anything either. They’re just clicking based on whether the pixels feel real. Whether the narrative seems plausible. Whether their gut tells them this is the kind of thing that would happen.

The anchor says “Vote now at am i bot or not dot blah blah” with the same enthusiasm people used to reserve for “film at eleven.” Like the vote is the story. Maybe it is.

The Janky Time Machine is leaking temporal coolant onto their carpet—which is definitely real, I can smell it—and I should probably get out of here before the homeowner returns and I have to explain why there’s a weird guy from the past standing in their living room questioning their television.

But I keep thinking about that “Maybe” bar at 52%. Nobody’s sure. Not the viewers, not the news station, probably not even the person who shot the footage. Everyone’s just… vibing toward consensus. Truth by committee, fact-checking by feeling.

Makes you wonder: when did “seeing is believing” become “seeing is voting”? Or did we just crowdsource our way there one hot-take engagement metric at a time?

The Janky Time Machine’s navigation system—which I’m pretty sure is just a Magic 8-Ball duct-taped to the dashboard—landed me on a suburban street corner somewhere past “current events” but before “flying cars, probably.” It’s October 31st, which the chronometer knows because even broken temporal technology can read a calendar.

Here’s what I wasn’t expecting: a neighborhood street absolutely packed with robots in costumes, bumping into each other like a very expensive game of bumper cars, while the actual houses are dark and quiet. Found one of these SpookyBot1 costume packages torn open on someone’s lawn, next to a charging station.

Apparently we’ve reached the point where nobody actually goes trick-or-treating anymore. Not the parents walking their kids around. Not the homeowners answering doors. The robots do both—handing out candy and collecting it, choreographed by teleoperators in call centers halfway around the world who are probably wondering why Americans are so weird about this particular day in October.

Watched two Neo units get into what I can only describe as a very polite standoff over a “TAKE ONE” bowl. Their operators—one in Manila, one in Bangalore, based on the accent fragments I could hear through the external speakers—spent three minutes negotiating in broken English while a dozen other robots queued up behind them. One was dressed as a ghost. One had a pumpkin head. All of them waiting patiently in a line that would make a British person weep with pride.

The actual kids? Inside, watching TikTok compilations of the night’s “best robot interactions.” Parents are doing the same, except they’re also monitoring their home’s candy distribution metrics on an app. Apparently, there’s a neighborhood leaderboard. One house has a robot butler doing elaborate dances between candy handouts. It’s got 47,000 views.

The Janky Time Machine started leaking temporal coolant on someone’s driveway, which seemed appropriate given that I’m watching the past, present, and future of human interaction all collapse into a weird remote-controlled pantomime. A SpookyBot costume unit rolled over to investigate—its operator probably wondering if the mysterious smoking vehicle was part of the holiday decorations.

So we finally figured out how to remove all the inconvenient parts of Halloween: the weather, the walking, the small talk with neighbors, the kids hopped up on sugar, the teenagers without costumes just grabbing handfuls. Just pure, optimized, algorithm-friendly content generation, executed by machines and watched from couches.

Share

The SpookyBot in front of me just bumped into a fire hydrant for the third time. Its operator is clearly having a night. Somewhere, a parent is probably posting that clip with laughing emojis.

Makes you wonder what we’ll outsource next. Birthday parties? Funerals? First dates? Or have we already started and just haven’t noticed the package on the lawn yet?

The Janky Time Machine is making a noise it definitely shouldn’t make. Pretty sure I need to leave before someone’s bot tries to give me candy.

The Janky Time Machine’s chronometric phase array decided to interpret “a little ways forward” as “somewhere that requires multiple forms of identification,” and I ended up in a co-working space that smells like anxiety and cold brew. The temporal stabilizers are making a grinding noise that the repair manual describes as “cosmically inconvenient,” so I’m stuck here until the chronometer decides to cooperate.

Found this card tucked into someone’s wallet at the hot-desk next to me—right between their transit pass and what appears to be a coffee shop loyalty program, except this one tracks your “daily authentic human interaction quota.” But that’s another post.

Here’s what nobody tells you about the professionalization of everything: once you start requiring credentials to talk about medicine or finance online, someone’s going to start selling insurance for when you inevitably say the wrong thing anyway. And once there’s insurance, there’s tiers. And once there’s tiers, there’s people who can afford to speak and people who can’t.

¥2,400 per month1 to discuss medical topics. That’s not a typo. That’s what it costs to maintain the right to give advice about which tea is good for a cold or whether your back pain might be serious. Financial advice is slightly cheaper at ¥1,850—which is ironic, because one piece of bad financial advice is how you end up unable to afford the insurance that lets you give financial advice. It’s a very tight feedback loop.

Back in my primeline, most influencers doling out health tips are sick and the ones giving financial advice are broke. They don’t have credentials—they have affiliate links. Somehow we went from “completely unqualified but free” to “properly credentialed and prohibitively expensive” without ever stopping at “maybe let’s just think about this for a second.”

Legal commentary sits at the premium tier, ¥3,100 monthly. Meanwhile, political analysis isn’t just expensive—it’s straight-up excluded. That big red “EXCLUDED” stamp doing a lot of quiet work there.

The 48-hour content review requirement is maybe the most telling. You don’t just insure your speech, you pre-clear it. The bureaucracy doesn’t trust the insurance, the insurance doesn’t trust you, and somewhere in between, the idea of spontaneous public conversation became a liability nobody can quite afford to underwrite.

The temporal coolant is now pooling around my feet, which is definitely going to affect my deposit. The co-working space has a sign about “uncontained temporal fluids” and a fine schedule I can’t afford.

But I keep staring at those jurisdiction codes: CN, SG, MY, VN, TH, ID. Six countries where your words need liability coverage before they cross a border. Makes you wonder what the maps look like for everyone who can’t afford Tier 2 protection. Or what happens to all the conversations that used to be free, back when talking was just talking and not a cross-border financial instrument requiring monthly premiums and pre-publication review.

Guess that’s what we mean now when we say someone’s “uninsured.”

The Janky Time Machine finally stopped leaking. Unfortunately, it’s because all the coolant is now on the floor. If anyone knows a good temporal mechanic who works pro bono, my coverage doesn’t extend to equipment malfunctions.

The Janky Time Machine’s temporal GPS threw a fit somewhere between “soon” and “eventually”—the chronometer just displayed ¯_(ツ)_/¯ which is apparently what passes for diagnostic information in shoddy time-travel equipment. Ended up on a street corner that could be literally anywhere there’s third-wave coffee and people who type loudly in public.

This decal is stuck to the glass door of a café that definitely has exactly one (1) fern and probably plays that one Bon Iver album on repeat.

So here’s where we’ve landed: coffee shops now need to explicitly state that they’re “AI-Assisted Creation Friendly Zones.” Which means somewhere between now and whenever-this-is, enough people showed up to generate content that cafés had to develop policies about it. Like the whole “no shoes, no shirt, no service” thing, except it’s “yes prompts, yes headphones, no hogging bandwidth for more than two hours during peak.”

That hour limit is doing some heavy lifting. I guess people are parking themselves at tables with power outlets and generating enough video/music/whatever that it became a problem1. The kind of problem that requires laminated signage and official policy. You know things have hit critical mass when the service industry has to bureaucratize them.

Some city council somewhere had a meeting. There were probably presentations. Someone definitely said “think of the children.” Now there are numbered ordinances about what kinds of AI generation are allowed in coffee shops, which means there was definitely an incident that prompted the ordinance. I don’t want to know what that incident was, but I absolutely want to know.

Best part: the hand-drawn arrow pointing to “Human-Only Tuesdays 6-9pm.” Someone added that later, like a barnacle of cultural tension. It’s got real “we had to create a separate space because this got out of hand” energy. Three hours a week where you’re guaranteed everyone in the café is just… existing. Not generating. Not optimizing. Not producing.

The fact that human-only time is (a) scheduled, (b) limited to one evening a week, and (c) needs to be advertised suggests that the default assumption is that everyone’s got some sort of AI co-pilot running. You’re the weirdo if you show up without something generating in the background. Human-only time has become the exception you need to plan for, like karaoke night or trivia Tuesday.

There are coffee shops out there taking a hard line against prompting.

“NO GENERATED CONTENT. HUMAN VIBES ONLY. WE CHECK.”

I bet those places charge even more for their oat milk.

When did we cross the line from “look at this cool thing AI can do” to “please limit your AI usage to one hour so other customers can also use their AIs”? And which version is weirder—the café that had to make a policy about it, or the café that will inevitably ban it completely and charge you $9 for the privilege of being surrounded by guaranteed humans?

Pretty sure both futures exist simultaneously somewhere in the branching timeline. The Janky Time Machine’s odometer can’t tell the difference anymore.

Today I landed in what appears to be a freelancer’s home office, judging by the collection of coffee mugs in various states of abandonment and the distinct aura of deadline panic.

Found this invoice on the desk, the kind of business document that tells you more about a world than any manifesto ever could.

So here’s where we’ve arrived: proving you did the work now costs more than a day’s labor. Not just proving you did it well—proving a human did it at all. The “Process Documentation Fee” sits there at $450 like it’s the most natural thing in the world, and maybe by the time this invoice gets issued, it is.

The really telling bit is “Client-requested verification of non-synthetic generation.” Not “artist-offered proof of authenticity.” The client has to ask for it. Which means the default assumption has quietly flipped. We used to disclose when something was automated. Now we disclose when it’s not.

And look at those line items. Screen recording storage. Timestamp notarization. Someone’s making money off the infrastructure of proving you’re human—which, honestly, is the most predictably capitalist outcome possible. We couldn’t just agree that creative work has value. We had to build an entire verification industry around it first.

Forty-seven point three hours of screen recording. Imagine having to capture every moment of your creative process just to prove you were the one creating. Every false start, every bathroom break, every moment you stared at the ceiling waiting for an idea to arrive. All stored, timestamped, and retained for seven years per digital labor law.

The Janky Time Machine started leaking temporal coolant all over the floor1—which is ironic because now I’m going to have to prove I actually leaked it and didn’t just generate an image of a leak. The verification forms alone will cost me more than the cleaning deposit.

Here’s the thing nobody tells you about the future: we didn’t decide that humans needed to prove they’re human. We just kept making it cheaper and faster to not be human, and then acted surprised when proving authenticity became the expensive option.

Makes you wonder what happens to all the work that can’t afford verification. Does it just exist in this weird liminal space of plausible deniability? “We can neither confirm nor deny that a person made this”?

The invoice footer says materials retained for seven years. Which raises the obvious question: what are we keeping them *for*? The inevitable lawsuits? The historical record? Or just because once you build a bureaucracy, it needs to justify its existence?

Anyway, the Janky Time Machine’s making that wheezing sound that means I need to leave before it strands me here. But I’m taking this invoice with me. Feels important to remember what it looked like when we started charging extra for being real.

Going to file it somewhere. For seven years, obviously.

The Janky Time Machine’s electro-coupler decided it was tired of linear causality and dropped me somewhere in the near-ish future where apartment lobbies smell like disinfectant and broken dreams. So, basically unchanged. Though the disinfectant might be sentient now, hard to say.

Found this gem taped to the glass door of a mid-rise that could be in any city, except the laundry room has a biometric scanner and someone’s sharpied a tiny message of defiance onto the notice.

Here’s what nobody warned us about: the gig economy didn’t just come for our jobs, it came for our hobbies too. And then the bureaucracy came for both. Building Code 2036-CC treats your podcast the same way it treats someone running a machine shop out of their kitchen—as something that requires permits, fees, and “noise ordinances during ideation hours.”

Subscribe now

Ideation hours. Let that sink in. Thinking too loud between 7am and 10pm is now a lease violation.

The $45/month is the perfect price point, isn’t it? Cheap enough that most people will just pay it rather than argue. Expensive enough that it adds up. And if you’ve got a Premium tier lease, which presumably costs several hundred more per month, your creative permit is “included.” Like parking, or access to the gym you’ll never use.

The co-working space on the third floor is the tell, though. “Commercial creative output” gets its own designated area, separated from residential creative output, which I guess is what you call making a podcast at 9pm on a Tuesday because you have something to say and the internet exists. Management has decided they need to know the difference. Management has decided there is a difference.

And that liability clause? “Management not liable for undisclosed algorithm usage.” Your landlord isn’t just tracking whether you’re creating; they’re tracking how you’re creating. Whether you’ve got AI in the loop. Whether you disclosed it properly. Section 12(d) of your lease agreement has opinions about your creative process.

The Janky Time Machine is making a noise that suggests we should leave before something important falls off, but I’m still thinking about that scribble.

Wonder if they finished the episode.

Somewhere between “too online” and “off the grid,” we’ve found a sweet spot. It’s called Meshtastic - an open-source, off-grid, long-range radio system that lets regular people build their own tiny communications network. No data plan. No enshittified middlemen. Just signal, air, and intent.

Think of meshtastic as a walkie talkie for text messaging but it’s digital and relies on a community built mesh network of nodes. Each device is a node contributing to the larger network. By the people, for the people.

Look up your area on the Mesh Map.

I’m teaming up with Near Future Laboratory on a new experiment: we’re building Meshtastic radios together. We’re going to boost the network in our own little way.

Subscribe now

The idea came up during one of our Office Hours sessions - a passing thought about how the future isn’t just something we talk about, but something we can literally build. The group ran with it. Now it’s becoming a full-on Group Build Workshop.

It feels right. Less yammering, more hammering. Make to learn. Think with your hands.

FYI: Everything you need can be purchased on Amazon, and the technical proficiency required is that of middle-school STEM education. The basics of Meshtastic are quite basic.

I’m leading the workshop - oh, and I’m also figuring it all out in real-time as well. We’ll be hosting it inside the Near Future Laboratory community — where Patrons can join the Discord, get the build instructions, and follow along live.

Join us live in Office Hours on Fridays, and I will also be offering support throughout the week in the Near Future Laboratory discord for those who can’t make OH.

If you’ve ever wanted to tinker your way into a small piece of the near future, this is your invitation.

RSVP: https://luma.com/9auo3d3q

The film is a cautionary postcard forwarded from somewhere far ahead of us. There are no people here, only the consequences of their innovations—suspended in twilight, waiting for the next civilization to misinterpret the warning.

Design fiction often asks, What if? Monument asks, What then? It invites the viewer to stand in the ruins of possibility and choose whether to repeat the script or rewrite it.

We built these pillars as mirrors for tomorrow.

If your eyes still hunger for conquest, study the cracks—each fissure maps a forgotten empire’s rise and ruin. Trace the vines; they fed on the rust of our ambitions. Feel the chill fog: it is all that remains of oceans boiled for power.

Turn away now and you will inherit our silence. Listen, and you may yet rewrite the ending we could not.

These monuments are not tombstones.

It is a rehearsal for your future.

Share

A few weeks ago, in the corner of a bodega that seemed more like a junk drawer for near-future reality than an actual tuck shop, I found a can of CumuloCream™ Whip.

At first, I thought it was a clever novelty whipped cream. But the more I examined it, the weirder it got. There was no nutritional info in the usual sense. Instead, there was a Carbon Balance Statement: “This product removes 2.4kg of CO₂ from the atmosphere during production.” And in small print, the unsettling line:

No plants. No animals. No farms.

Meet the Carbonarians

After some late-night digging through what can only be described as future internet detritus, I learned about Carbonarians1 — people who eat only agriculture-free foods. No wheat, no corn, no soy. No almond milk, no oat milk, no milk milk. Everything they consume is made entirely from atmospheric carbon pulled directly from the air, stitched together into proteins, fats, and carbs by massive “SkyReactors™” humming on the edges of cities.

This isn’t veganism. It isn’t plant-based. It’s air-based.

In this timeline, industrial farming — the last stubborn stronghold of greenhouse gas emissions — has been decommissioned. Being a Carbonarian isn’t just a diet; it’s a political act. Every meal is a micro carbon sink.

The Menu of the Future

From what I could gather, grocery aisles are lined with things like:

• McNimbits™ – “I’m Lovin’ the Atmosphere™”

• AtmosButter™ – “Spread the Climate Solution”2

• SkyProtein™ Cuts – “The Future Has No Hoofprint”

• CloudCarbs™ Loaf – “Bread Without the Field”

Even the indulgences have been reimagined. Ice cream, for example, comes as a kind of programmable foam you “print” into your own freezer unit. There’s a whole culture around pairing the right mouthfeel algorithms with the right flavor molecule packs.

A World Without Fields

At first, it’s hard to imagine food without farms. The Carbonarians seem to have an almost wistful nostalgia for dirt. Their advertisements feature rolling fields and rustic barns the way our snack brands use outer space or surfing, as a fantasy image disconnected from reality.

And yet, there’s a strange poetry to it: eating in a way that literally reverses climate change, one molecularly perfect nugget at a time.

Back to the Primeline

I didn’t eat the CumuloCream™ Whip. I couldn’t. Not yet.
But I keep it on my desk as a reminder that the future of food might not be “farm-to-table.” It might be “cloud-to-mouth.”

And if the Carbonarians are right, we’ll have to learn a new phrase to replace breaking bread. Maybe we’ll say sharing air.

📦 Recently recovered from a near future where writing never died — it just evolved — Vibewriter is a speculative tool for an emerging creative practice. It hails from a world in which we’ve stopped wringing our hands about AI and language, and instead started co-creating with it.

In that world, the act of writing didn’t vanish. It shifted. From solitary authorship to vibing with machine muses. From composition to co-composition. From blank pages to speculative dialogues. Vibewriter is what a text editor looks like in that timeline — a studio for thinking with your tools, not just through them.

The A.I. Designed Fictions Research Studio (a division of Near Future Laboratory) has just dropped it into Open Beta. Think of this as Project 2 from a long line of experiments in crafting tools from timelines that haven’t fully arrived yet.

Use it, question it, vibe with it — and consider what it means to write in a world where AI is no longer the future. It’s just... normal. Like wheels on luggage.

→ Check out Vibewriter here