All Articles
Longreads
Mapping the Next Era of the Internet
Everyone agrees we’re entering a new era — but it’s going in a different direction than we might naturally think.
My identical twin and I have built on the Internet for over 30 years. (In twin-years, that’s over 60 years of experience.)
We’ve seen a few things! Perhaps the one constant has been that the Next-Big-Internet-Thing is ever changing (and we jump on it too quickly).
But in 2023, we had a new experience: we got stuck.
We’d started building Pathwright 3 with a clearer vision, a more skilled team, and more engaged customers than ever. But our vision for what to build began confusingly shifting. It was as if the tide of the Internet was sliding out under our feet.
Was it the post-pandemic digital reset? A tidal wave of AI? A midlife crisis?
And then we had an epiphany.
My focus at Pathwright is mapping our big-picture strategy. ✎ (I’ve got a thing for maps.)
In this post, I’ll sketch a new map that I hope will give you a calmer, clearer big picture of how the Internet’s evolving and how we’re adapting to it.
Internet 1.0 ∙ Mid-’90s - 2010
Let’s begin at the Internet’s mainstream debut — the “You’ve got mail” era.
Looking back, this era was Weird, Indie, and Local.
1.0 • Weird
Remember GeoCities? There were no rules for making a webpage, so why not use wild tiled backgrounds, looping hamster GIFs, embossed buttons, and drop in the all-important visitor counter?
When we first logged onto the Internet, Mark and I were a bit weird too:
In case you can't tell by this picture, we were homeschooled.
It was 1996. We were thirteen, obsessed with our computer, and sweating over HTML for Dummies. Our mom — who had no idea what the Internet was — thought whatever we dummies were up to seemed sciency and might be a big thing one day. So she assigned us an hour per school day to learn HTML, unwittingly creating what was possibly the first high school web-design class.
During that hour — and in every spare moment — we built GeoCities sites lampooning friends, wacky webpages full of <blink/> tags, and earnestly cringey Flash intros. (Remember Flash intros? They took ages to load on dial-up and then 30+ seconds to play before the website even loaded.)
Somehow, we got a few of our parents’ friends to pay us $500 a pop to create websites (and even more to add a Flash intro).
Building on Internet 1.0 made us feel like superheroes — we could demolish two large pizzas in one sitting and build any webpage we imagined and get paid for it.
1.0 • Indie
In the early web era, everyone in tech felt like a superhero. Anyone with a dream, a modem, and a garage could disrupt an industry.
Sergei and Larry in a rented garage where they created what became Google
The myth of the garage spread everywhere dial-up lines reached—even our family den in Greenville, SC. A local client connected us to his San Francisco-based startup, and suddenly, we were earning $20+/hr coding flash intros and elaborate UIs on top of ColdFusion code.
Even after that project imploded with the Dot-Com crash, the Internet revolution quietly continued in San Francisco without us. (But we still managed to pay our way through college by building software for our university and other clients.)
Then came the iPhone. Solo developers made millions on apps like Koi Pond, iBeer, and iFart. Social networks amplified ground-up, local movements like Occupy Wall Street, the Arab Spring, and later the message of HOPE, landing Obama in the White House.
In this era of Internet-fueled optimism, it felt like a handful of people could build something they loved and change the world.
By 2012, Mark and I launched our moonshot— Pathwright, the first platform to enable anyone to create and sell online courses under their brand.
1.0 • Local
The small, independent builders of the early Internet solved local problems in local clusters. But a few accidentally stumbled into solutions with universal appeal:
Mark Zuckerberg infamously built “FaceMash” from his Harvard dorm room to rate how hot his fellow coeds were before pivoting it into “The Facebook.”
A one-off web page listing a spare room and an air mattress to attendees of an overbooked SF tech conference became Airbnb.
My favorite example: Sal Khan used his tablet and a new website called “YouTube” to share an annotated video explaining math concepts to his cousin, Nadia. (It turns out Nadia wasn’t the only viewer… more on that later.)
Pathwright also began as a solution for one client who was in dire need of an alternative to Moodle for offering online courses.
Pathwright began as a simple, nicer replacement for Moodle.
Pathwright centralized Moodle’s chaos of docs and links into a unified set of steps anyone could follow. This pattern worked so well for our client that it spread to ten others before we left beta.
A Tide Change
In hindsight, that core benefit of Pathwright — centralizing decentralized content — was a micro-example of a larger pattern sweeping across the Internet:
“The Internet had shifted from a decentralized era to a centralized era we’ll call “Internet 2.0.”
Facebook founder Mark Zuckerberg meets President Barack Obama in 2014. White House
Internet 2.0 ∙ ~2010 - ~2023
Internet 2.0 emerged from 3 reactionary centralizing forces: Scale, Uniformity, and Conformity.
Local → 2.0 • Scale
Apps that worked locally and happened to solve universal problems scaled everywhere the Internet reached at an unmatched pace:
Facebook spread like a pandemic from Harvard, to other campuses, to 40% of the world’s population, including — perhaps especially — your mother and grandmother.
Sal Khan’s math videos became ✎ Khan Academy — the largest educational non-profit in the world.
Google graduated from a nerdy academic project targeting enterprises to become the Internet’s front page.
Meanwhile, Pathwright slowly grew from a side project to a full-time gig, we felt pressure to scale fast or die. Well-meaning investors urged us to pour venture-capital-gasoline on our growth or risk being relegated to a “lifestyle business.”
A brief aside:
Every Business Is a Lifestyle Business
Thankfully, our independent roots and local advisors helped us see that every business comes with a lifestyle:
Hustle for a 10X return and exit in 5 years richer, against all odds, but with little control.
Partner with customers to grow steadily with independence and creative freedom.
We chose independence and haven’t looked back.
Weird → 2.0 • Uniformity
Patterns that scale mature into standards:
Weird personal sites/MySpace → Boxed-in Facebook/LinkedIn profiles.
Custom courseware/hosted Moodle → SaaS education platforms, course marketplaces, and MOOCs.
Personal blogs via RSS → Substack & Twitter.
By 2020, if you squinted a little, every app or website in a given category looked interchangeable.
For instance, virtually every online course platform uses a left sidebar and right-side content (a trend we intentionally dodged):
In 2017, we shipped Pathwright 2, a scaled—up version of PW-1. We maintained our weird, indie roots (like avoiding the sidebar trend) but also employed a more usable, flat 2.0-era design.
Indie → 2.0 • Conformity
The few giants that dominated the late-stage Internet 2.0 era forced it past uniformity into conformity.
Some conformity to sane governance is good — the digital wild west needed taming. But there’s a subtler, creeping conformity that’s clogged the Internet’s arteries:
SEO Slop: As Google became the gatekeeper of the Internet, Internet content once made for humans got littered with keywords, mind-numbing filler, and cross-links to serve Google’s algorithm.
Social Media Addiction: As influence spread with seemingly no ceiling on social media, content trended towards superficial clickbait tuned to go viral with little care for what it does to our well-being. ✎ Even social media’s UI itself is bad for us.
By the early 2020s, we’d grown disenchanted with Internet 2.0. In pursuit of scale, it had become a sluggish, uniform behemoth peddling endless, ad-ridden feeds of clickbait slop that taxed our time and energy.
Ironically, what is at once the greatest library of learning opportunity in history is now also ✎ the hardest place to deeply learn (or even focus).
We missed the simpler, weirder, more personal days of the Internet.
Even Grandma’s handwritten recipe cards have a nostalgic glow compared to the endless SEO-bait-filled recipe blogs plastered with ads.
And it wasn’t just us. A broader Internet 2.0 backlash had begun.
An unexpected, but familiar, pivot
Our early designs for Pathwright 3 were naturally a bigger, better Pathwright 2. Because progress looks like up and to the right, right?
But the further we got, the more we felt a pull to go smaller. To make individually personalized, adaptive Paths possible. We began experimenting with weirder interfaces, AI, and local data structures that don’t fit as a progression from a Pathwright 2.0 — or Internet 2.0 — paradigm.
And then we realized we’d seen this all before.
None of the above are real—all were generated. But they sure drove engagement.
Internet 3.0
The centralized Internet 2.0 wasn’t progressing up and rightward towards greater scale, uniformity, or conformity; rather, it was pivoting back into a new decentralized era.
“History doesn’t repeat, but it rhymes.”
— Attributed to Mark Twain
Human progress isn’t linear — it cycles between centralization and decentralization like a spiral.
Internet 3.0 is emerging now atop Internet 2.0 and in opposition to it. If you look closely, you can see the Internet slowly transforming:
Uniformity → 3.0 New Weird
✎ NFT Bored Apes sold at Sotheby’s for $24 million and crashed in value by 90%.
Non-developers are ✎ vibe-coding apps.
There’s ✎ a gold rush of funding for anything AI (with a bonus for crypto mix-ins).
Web pages are ✎ getting weird again.
✎ Spatial computing is finally here (sort of), and a lucky few are designing and experiencing digital interactions that were only seen in sci-fi not long ago.
There’s a renaissance of weird and “woah” if you know where to look.
Scale → 3.0 {New Local}
Blockchain has already experienced its own Dot-Com-style boom and bust, but is ✎ quietly decentralizingcommerce and contracts, laying the groundwork for another level of decentralization.
Many people — especially the younger generation — are ✎ abandoning public social feeds to bots and news and returning to AOL-IM-style private group chats on Telegram, Discord, or WhatsApp for real conversations.
Even Google search has increasingly been replaced by private, individualized chats with Claude or ChatGPT. Things are so localized that it’s hard to know ✎ what's happening on the Internet anymore.
What’s more, ✎ the next wave of innovation is clustering in real places again:
“For something so up in the cloud, AI is a very in-person industry,...People are going to parties, hacker houses, happy hours. It’s where they all mingle and exchange information.”
— Jasmine Sun, quoted in The New York Times, August 4, 2025
In education, we’re seeing a similar shift to local with post-pandemic Zoom-fatigue pushing people back to more real life, group environments.
Conformity → 3.0 New ⚑ Indie
The current revolution in AI was sparked ✎ by a paperwritten by a team of 8 people at Google. Individuals and small groups like these are making the biggest dent in blockchain, decentralized notes apps, and new local-first development frameworks.
Meanwhile,✎ open source is making a comeback, particularly with new on-device (very local), open-weight AI models.
In turn, AI now provides any creative, ambitious person with a new kind of “garage” from which to build a massive online multiplayer game, make a movie, or generate a bespoke app.
Now, anyone with an Internet connection + AI can and will do things we could only dream of just a few years ago.
Pivoting Back to the Future
The slow epiphany of where the Internet — and our hopes for Pathwright’s next version — were headed back to led to a pivot.
We refocused the majority of our team on making the current version of Pathwright the best Internet 2.0 product it can be — simpler, more modern, and better at what it does best. (The centralized Internet 2.0 isn’t going anywhere — rather, it is graduating to become the new infrastructure on which 3.0 products will be built.)
Last year, I posted an update to our “Backstage Path” letting our customers know of a change of plans for “Pathwright 3”
Meanwhile, a smaller Labs team is building a new product for the next era that expands what’s possible for learning and teaching down to the most decentralized scale: one person.
This dual focus has given us increased clarity and focus on both fronts, and our only regret is that we didn’t realize where we were headed sooner.
Bravely Entering a New, Old World.
Even if progress in AI or blockchain were paused today, I’d bet it would take close to a decade of local, weird experiments to find new patterns that “just work.”
It’s an exciting, chaotic, figuring-it-out transitional era. There are no predictive paths, playbooks, or timelines. But here are a few things we’ve been relearning, having been here before:
This is an era to...
Be Weird: Hone in on the unique personality and value you bring; The era of “looking big” on LinkedIn, or playing by centralized-era playbooks, is over. People are looking for what’s next in opposition to the status quo — being unique and authentic to your personality and mission is square one.
Focus on {Local}: Era-sized transitions are the perfect time to hone in on solving specific problems for individual people you care about (ideally, starting with yourself). For example, until we can confidently demonstrate significant ongoing benefits of a tech like AI or crypto for one learner, it’s premature to dream about scaling it to teach hundreds or thousands.
Stay ⚑ Indie: The venture-fueled Internet 2.0 growth paths are no longer safe, if they ever were. In a transitional, decentralized era, nothing is worth more than independence — the psychological and economic flexibility to adapt on your terms, making bets where you see fit (and no investor would ever spot or allow). In an era requiring experimentation, adaptation true independence looks like investing in work you’re satisfied with, even if it only solves a problem for you and a handful of others. If you’ve got a small team (or just yourself), good for you —we’re reentering an era where a two-pizza-sized team is a massive advantage.
What do you think?
We’ll share more about what we’re learning about this new era as we build.
In the meantime, I’d love to hear your thoughts and tips. What are you seeing and learning as we enter this new era?
