A memo from the future

It’s 2069. Congratulations. We’re still alive.

If you time traveled here straight from 2025, you wouldn’t recognize the world. You’d find it as radically different as your great-grandfather would find 2025 if he could have teleported to it from the 1930s.

As blogger Tim Urban wrote in “Meet Your Ancestors,” your great‑great‑great‑grandfather in the 1800s lived most of his life “without running water or electricity.” Go back 18 generations to the Viking era, and your great²⁰-grandfather was crusading, dodging the plague, and torturing and/or being tortured — in the words of Urban, he’d be “blown away by the ease of your current pussy existence.”

The people of 2069 look at you like you have two heads when you tell them that cancer was once a leading cause of death. That’s because they can take a pill that kills cancer so fast, they’re back to work the day after treatment.

Your slow, pondering, linear thinking drives them nuts too. In 2025, you can only think about one thing at a time because you only have neurons in your skull. But flash forward to 2069 and a huge population of early adopters has computers in their heads. AI didn’t make people obsolete — it moved inside and merged with us.

Intelligence infuses everything, and once-dumb objects have woken up.

The latest neural laces make it easy to talk to friends, family, and co-workers — communication feels like the magic of telepathy. These devices also monitor your health, and let you know to get to a doctor right now because you’re going to go face down into your oatmeal from a heart attack in a few weeks if you don’t. When you see the doctor, he gives you a pill equipped with tiny micro-machines that eat through the arterial plaque. A second pill uses mRNA to reprogram the genes that regulate how your body produces cholesterol.

Intelligence infuses everything, and once-dumb objects have woken up. Your shirt is gossiping with your shades, which are bantering with your partner’s pearl earrings as she lands in São Paulo for a conference. The streetlights are whispering to the asphalt about black ice forming in the shadow of the stadium. Containers hum lullabies to their cargo as they cross the Pacific, while nanosats in low Earth orbit monitor the economy in real time as a global swarm intelligence peppered across the sky.

Everything is alive. Smart bandages watch your wounds and ping you if the micro-infection they’ve detected isn’t clearing. The embedded pharma patch changes tactics mid-fight, switching antibiotics because the bugs called in reinforcements. Your pill bottle knows the medicine it contains interacts badly with alcohol — when you try to open it at 2 a.m. after a night out, it flashes a stern “don’t you dare until you hydrate and sleep” message, gleaning from an implant that your blood-alcohol level is too high.

Self-driving trucks without cabs scream along dedicated freightways in America, Europe, India, and China at 200 miles per hour. They’re stripped-down electric sleds with wheels, hauling standardized smart containers that know every object inside them — each piece of freight was tagged at birth, logged to the world’s distributed databases, and is now tracked by a whispering mesh of nanosats. As soon as a dish slides out of a fully automated dark factory’s industrial 3D printers, it’s in the ledger and on its way to you. When you drop it, it orders its own replacement.

If pirates angle for a container ship, it yaws and jukes like a prizefighter, while its containers beam distress signals straight up to the eyes in the sky. A squadron of unmanned defense drones then sprint from a nearby ship to bring the pirates down with non-lethal attacks.

As you stand outside the legendary Berghain club in midwinter Berlin, your jacket tightens its microthreads to keep you warm. The bouncer finally lets you in with a nod, and the jacket loosens the instant you cross the club’s threshold. Layering is for the old world. The robot coat check — an elegant spool of floating hooks — irises your ID and sends the claim code to your info-necklace. You dance ‘til Sunday afternoon while a million invisible AIs keep the city purring.

The industrialized future

The world of 2069 is not a world without humans. It’s a world where humans have more reach, more speed, and more leverage than any previous generation — and it’s starting to take shape in 2025.

We are currently entering the age of industrialized everything. We’ve already seen Act I. A decade ago, tools started crushing cost and time across the creative and coding stack, and output exploded. In entertainment alone, teams of 50 can now accomplish feats that once required 500 people — and instead of fewer jobs, we’re getting more titles, riskier greenlights, and a far wider cultural palette. Cheaper creation isn’t killing careers. It’s multiplying them.

Fifty years from now, that same cost collapse will reverberate across every industry.

The robotics flywheel

Every generational platform has a self-accelerator baked in. For software, it was app stores and compounding libraries. For robotics, it’s the loop that starts the second a new robot touches the world. Action generates data, data trains models, and models make robots smarter, more capable, and able to handle more edge cases. Cheaper robots multiply deployments, and deployments generate more data.

Once one robot knows something, they all do. Unlike people, who pass around lossy learning with sound and writing, robots download updates to their mental “weights” and patch them in. Download done, patch applied. Your dog-walking robot can now walk your kids to school. An update last week taught it to do the dishes and improved its deep-cleaning skills.

Around and around it goes, accelerating every quarter. Point that loop at anything repetitive or physically gnarly — logistics, construction, agriculture, surgery, disaster response, etc. — and it gets progressively more automated and more resilient.

We’ve learned to build the loop on purpose and trust it. Assemble the compute and the machines. Align the capital. Plug in the intelligence. Set the incentives, and then get out of the way while the machine compounds.

Open exchange and innovation lead to more surpluses and more openness. That’s the meta‑flywheel.

There’s a bigger flywheel humming under all of this: openness. Golden ages of the past didn’t happen because someone found the perfect slogan. They happened because cities opened the doors to people, to trade, to taboo‑breaking ideas, to technology, and to progress — and then kept them open against all odds.

Swedish author and historian Johan Norberg nails the pattern in his book “Peak Human”: 

Athens ran on porous borders and direct democracy. Rome brought in bigger and bigger waves of people and land, then wove it all together with aqueducts and paved half the world — goods and knowledge fanned outwards and routed back to Rome in a self-reinforcing loop of knowledge and goods. Song China protected property, printed money and books, and hired state administrators on merit instead of nepotism and family bloodlines. It then watched cities explode with invention until reactionaries pulled the plug.

Open exchange and innovation lead to more surpluses and more openness. That’s the meta‑flywheel. Technology lowers friction; openness accelerates it. More compute, more energy, more biotech — plus more visas, more trade lanes, and more intellectual free-fire zones — equals more people doing weirder, better work. We didn’t “lose jobs” when 99% of us stopped farming. We gained time, options, and whole new categories of human purpose. We’ve destroyed thousands of jobs throughout history, and each time, we’ve replaced them with better ones. That’s called progress.

More tech and more openness is a rising tide that lifts people faster than less tech and less openness. Close down, and you shrink. Open up, and you compound.

The engines of abundance

If you want a future worth living in, you need engines that make scarcity disappear. 

Start with energy. Fusion won’t arrive as a single miracle box that flips overnight from lab toy to grid savior. It’ll arrive as a long chain of solved engineering problems: confinement tricks that actually scale, tritium production that doesn’t tank the economics, heat extraction that doesn’t cook your walls, and materials that laugh at neutron abuse. The real tell is capital: When Microsoft signs a fusion power agreement with Sam Altman-backed startup Helion long before “grid‑scale” is table stakes, it’s a vote of confidence in big changes to the grid coming soon.

By mid‑century, orbital solar power will stop sounding like a Bond villain plot and start looking like smart grid glue. Cheap launches and city‑scale rectennas will mean you don’t just power Manhattan from space — you power the 3 a.m. robotic factory surge without spinning up a gas peaker. Early SpaceX Starship‑era pushes into making the space race cheap and repeatable will compound, and as launches get cheaper and faster, we’ll put more powerful machines into space: power stations, drilling machines, robotic launch platforms that can shoot off more rockets, and solar sail ships that go even deeper into the cold dark of space. Early demos, like Caltech’s Space Solar Power Project, have already proven that power-beaming from orbit is possible, and not sci‑fi.

Compute is the second engine. The 2020s will go down as the decade we learned to feed the model beast, with H100s, B200s, and custom accelerators stacked to the ceiling. The 2030s will be when light takes over. Photonic interconnects and photonic compute will torch inter‑node latency and kill the power wall for specific workloads. Companies like Lightmatter are already showing the way with light-speed chips that make electronic buses look like flocks of carrier pigeons.

After that is when wetware takes over. While companies like Meta need to build electricity-guzzling data centers the size of Manhattan to support their AI efforts, the human brain can run on a bag of potato chips. Merge biotech and computers, and you slash your processing power requirements. Imagine hybrid supercomputers made of hundreds of trillions of human neurons running tomorrow’s superintelligent systems. This living wetware could power our grids and our economies at a superhuman level. Australia-based biotech lab Cortical is leading the way, growing real neurons on computer chips, and a pioneering research group at Indiana University Bloomington has demonstrated that neurons connected to chips can process and remember information. They can even perform basic speech recognition.

Resistance to gene editing will fade as people experience its benefits for themselves.

Biotech is the third engine. AlphaFold turned protein structure prediction from a years-long slog into a simple database search. AlphaFold 3 then enabled multi‑molecule predictions. This will crack drug design and synthetic biology wide open. CRISPR will move from the lab to the clinic with first‑in‑class approvals, and by 2069, personalized gene therapies will be commonplace. Instead of prescribing treatments that might work based on your blood work, doctors will use digital twins of your physiology to test diets, drugs, and other therapies in silico before ever needling your skin.

The action will be in programmable cell factories, on‑demand biologics, and whole‑body rejuvenation protocols that reset epigenetic drift without blowing up your risk profile. The FDA recently approved the first CRISPR gene therapy. In vivo edits will follow. As far back as 2021, the New England Journal of Medicine was reporting on experiments to edit away the life-threatening disease transthyretin amyloidosis right in the womb. The understanding is there. As the cultural taboo of gene editing falls away, we’ll get to refining the technology.

And why will resistance to gene editing fade? Because people will experience its benefits for themselves. That’s when they’ll realize that the doomsayers and the rabid anti-tech fanatics are disconnected from the real world. They see the truth with their own eyes. The horrific and fatal disease that their child would have been born with will become totally preventable. People everywhere will embrace a better life for themselves and their kids.

Materials form the fourth flywheel pillar. Discovering a new material used to be a crapshoot — researchers would try a thousand powders and pray that one would be remarkable. Then machine learning put a rocket under materials science. DeepMind used large-scale parallel AI searches to map out millions of potential new materials — a digital treasure map for chemists. By 2069, finding a desired material will be as simple as typing a search query: “Give me a room‑temperature flexible conductor with these tensile and heat characteristics” — and the pipeline will return a recipe.

Life and work in 2069

Morning light spills through glass that isn’t glass. Your home’s “skin” tracks the sun and quietly arbitrages electricity with the neighborhood microgrid. Surfaces remember your comfort band and your guests’ quirks. The shower preheats itself while your personal AI trades five minutes of peak power for cheap overnight storage behind the scenes.

Your med‑stack hums in the background after you use your smart toilet that just arrived from Japan. It runs prevention routines by default. Agents watch biomarkers for drift and nudge you toward micro‑interventions that never become macro‑crises. A full‑body super-MRI comes with your health plan subscription — now you catch the tumor when it’s 12 cells big, not 12 billion. When you need a drug, your digital twin co‑designs it, and the local biopharmacy prints it before lunch.

Mobility is orchestrated rather than chaotic. Arguing about “full self‑driving” in the 2060s reads like arguing about whether to take a horse or a car in 2025. Everything is fully self-driving. People once refused to get in a car driven by a machine. Now it’s the opposite. They don’t want to get in a car driven by a human who can speed, talk on their phone, or get distracted — those flesh-and-blood drivers are 1,000 times more likely to cause accidents. Trucks, buses, pods, bikes — every vehicle negotiates with everything else, second by second, the grid making them smarter. Zipline-like drones with no visible rotors thread the skyline, transporting groceries, meds, and hot tacos.

Housing is printed rather than poured and erected. Swarms of construction robots lay bones. Humanoids with fingers set the artistic flourishes and finishes. Material‑handling bots do the heavy lifting. ICON’s early 3D‑printed neighborhoods were the opening riff. NASA even hired them to figure out how to print on the moon.

Your coworkers are an isomorphic swarm of specialists running locally for privacy and in the cloud for scale.

Work is less “job,” more “quests.” A dozen micro‑firms — half human, half AI agents — stitch together for a six‑week push. Then they dissolve and recombine. Your cofounder is a swarm, your board meeting is a simulation, and the company is wherever you are.

We finally built tutors that actually teach. A great teacher watches “tape” of you and figures out your tells, your gaps, and your blind spots so that it can bring you just the right drill at just the right time. Khanmigo was a charming early attempt, but it looks like caveman tech next to a real “Diamond Age”-level tutor that knows you and your children better than you know yourselves — it’s a perpetually friendly and patient companion who is leveling you up your whole life.

Your coworkers are an isomorphic swarm of specialists — design, legal, capital strategy, ops, etc. — running locally for privacy and in the cloud for scale. You don’t “use software.” You hire a team of tireless digital weirdos who never sleep to build what you need on the fly in a few days. The “company of one” stopped being the subject of Substack think pieces and became a funding category. There are now multiple one- and two-person companies worth billions. With an AI crew and the global agent economy, a single founder can get hardware created in Shenzhen on Monday, run clinical trials in Nairobi on Tuesday, and close distributors in São Paulo by Friday — without leaving their kitchen table.

The last era of phones looks as quaint to the people of 2069 as rotary dials do to us today. Computers got out of the rectangle and into our senses. Glasses became the default display: lightweight, beautiful, boring. They translate a multi-language café conversation in real time and summon a rideshare with a glance. Ray‑Ban Meta glasses were the appetizer.

Audio became the new command line. Constant, contextual conversations with your agents beat tapping glass. Your AI knows when you’re heads‑down and don’t have time to read text. An invisible ambient assistant is just there, like a soft presence. Neural I/O will graduate from science project to last‑mile accessibility and high‑end pro tools. Neuralink’s first implants are clumsy miracles; by the 2060s, rigs are invisible, safe, and boring.

The builder’s playbook

If you’re building in this world, here’s the condensed playbook:

Pick a flywheel — robotics, biotech, materials, AI, etc. — anywhere where doing the work generates the data that makes you better at the work. Design for the loop from day one. Own a data moat that isn’t just a pile of files. Own semantics, simulators, real‑world feedback channels.

Put atoms in the loop. If your output sits purely on a screen, you’re in a knife fight with a hundred thousand founders. Move matter. Deliver joy to a hand, not a timeline.

Monetize the mess. Infrastructure companies mint money by cleaning up chaos: observability for robots, insurance for swarms, debugging stacks for embodied AI, test farms for edge models. Be the boring backbone and cash checks forever. Sell outcomes, not software.

“We reduce unplanned downtime by 67% for offshore wind” beats “We have a dashboard.” Tie price to value. Put skin in the game.

The future is not a committee meeting. It’s a hands-on tinkerer’s lab where you learn as you go.

Design for agents as well as humans.

In 2069, your customer is a person plus their AI cohort. Give agents purchasing authority within constraints. They will use it. Then ship the weird. When the price of failure craters, orthodoxy is for cowards. The market is starving for flavors it’s never tasted. Serve it something dangerous and delicious.

The next 50 years won’t be smooth. Nothing real ever is. But the toolkit to build a better world has never been richer: abundant energy, fusion power, wetware, dark factories, robotics flywheels, mRNA and CRISPR, agent swarms that turn wishes into working systems, cities that think, medicine that prevents more than it treats, education that finally cares how you learn, and rockets that turn low Earth orbit into a suburb.

This is a love letter to the makers. The future is not a committee meeting. It’s a hands-on tinkerer’s lab where you learn as you go. It’s the patient AI tutor looking over your shoulder and helping you get better faster. It’s an open workshop.

Go big. Go bold. Ship the impossible.

We’ll meet back here in 2069 and swap war stories.