I’m still in awe of how much data we create — and how fast that number moves.
Back in 2011, I was amazed that users created 600+ new videos and 60 new blog posts every minute … and I was talking about how many people I saw at the mall (though it was decreasing …)
The Internet is both timeless and timely in an interesting way. While what’s popular seems to be ever-changing, what it does (and what we are capable of doing with it) continues to grow exponentially. Ultimately, the Internet is the digital town square of a global village, where all types of participants gather.
In 2011, I first wrote about what happens on the Internet in 60 seconds.
Each time I write the article, I’m in awe of the amount of data we create and how much it has grown. For example, looking back to 2011, I was amazed that users created 600+ new videos and 60 new blog posts each minute. Those numbers seem quaint compared to current figures.
Today, the Internet reaches approximately 6.2 billion people. Most of them also use social media.
To add some more perspective,
In 2008, 1.4 billion people were online; by 2015, that number was 3 billion. Now, that number has doubled again.
In 2008, Facebook had only 80 million users, and Twitter (now X) had 2 million.
In 2008, there were 250 million smartphones; now there are over 7 billion!
It is mind-blowing to consider what happens on the Internet every minute today.
In 2023, the world created approximately 120 zettabytes of data, which breaks down to about 337,000 petabytesper day. Broken down further, that’s more than 15 Terabytes of new data created per person… now scale that with the rapid growth of Generative AI, and even another billion internet users.
A growing share of what fills an internet minute is now generated by systems — models answering models, agents writing to other agents, sensors reporting to services nobody reads.
Globally, generative AI platforms receive roughly 2.5 billion prompts and serve between 115 million and 600 million daily active users. Over 1 billion people use AI tools each month, and ChatGPT alone reaches over 900 million weekly users.
Can you imagine how much data that is a day? Can you imagine how much more data will be created in five years?
It’s still too early to tell whether the scale will be exponential … or logarithmic. Regardless, I think we’re moving to a post-human tipping point where technology starts to drive more of what happens on the internet (as more devices and digital WHOs create and share data, it’s hard to fathom the ramifications and the sheer volume of data), and beyond that … we’re entering an era where an increasing amount of data will be generated in space.
So, while a lot already happens on the internet every minute… I believe the safe bet is that it will move even faster next year… even if human usage stays exactly the same.
To some, new technology is a good thing. To others, less is more. It’s been interesting watching many of my formerly tech-savvy peers shy away from AI.
But it’s a tale as old as time.
Most people simply “tolerate” technology transitions; some people drive them, and others crave them and use them as a catalyst for growth or strategic advantage.
Everett Rogers published Diffusion of Innovations in 1962, studying how Iowa farmers took up hybrid seed corn. He found five groups, in stable proportions: Innovators (2.5%), Early Adopters (13.5%), the Early Majority (34%), the Late Majority (34%), and Laggards (16%).
Thirty years later, Geoffrey Moore added the part that actually hurts. In Crossing the Chasm, he argued there’s a gap between the early adopters and the early majority — and that most technologies die in it. Not because they don’t work. Because the people on the far side want something the enthusiasts never needed: proof, references, and someone else to go first.
Seed corn in 1962. Personal computers in 1985. The web in 1997. Smartphones in 2009. AI right now.
The description begins with resistance and progresses towards compulsion. Reversing this sequence allows us to illustrate the innovation adoption process.
Here is a visualization of the innovation adoption model and market share.
In the image above, the blue line represents consumer adoption, while the yellow line represents market share.
As you can see, only 2.5% of the population drives innovation (or adopts it early enough to help drive the Alpha & Beta versions of emerging technologies). 13.5% make up the Early adopters, who help get it ready for the mainstream. Then, the early and late majorities are the groups that ultimately consume (or use) the mature product. Meanwhile, Laggards are often forced, kicking and screaming, into “new” technologies as the early adopters are well on their way to subsequent iterations.
Sixty years of data, and the proportions barely move. That’s the finding. The curve isn’t really about technology at all — it’s a snapshot of how people handle uncertainty, and it happens to get re-photographed every time something new shows up.
Times Are Changing …
Here’s where the analogy breaks, though — and it’s worth saying plainly. Every prior transition gave the late majority time. You could ignore the web for four years and catch up in one. The compounding was slow enough to forgive a late start. That’s the part I’m not sure survives this time, and it’s why the peers I opened with are on my mind.
Even if you are not an innovator, here are a few Innovator Mindsets that I find useful.
You Believe There’s A Better Way
Wherever you are, you know that there is a best next step, and you are eager to find it and take it.
You recognize that the opportunity for more (or better) often lies just beyond the constraints or problems of the current way.
The bigger the future, the more your efforts fuel it. When initial excitement fades, understanding what the bigger future can bring helps you power through.
You Are Comfortable Being Uncomfortable
You understand that Pioneers sometimes take arrows in the back.
When creating a new reality, you expect some resistance as a result of the law of averages. Escaping the status quo takes a lot of momentum, but it’s worth it.
You recognize when victory is near. In a quirk of human nature, too many people quit just before they would have won. Don’t make that mistake.
You Know Where You’re Going, Even If You Are Not Sure How You’re Going To Get There
Your goal should be your North Star. A clear direction is essential to ensure that activity leads to progress.
Measure progress and momentum rather than the distance from your goal.
It is easier to course-correct while in motion.
If you’re too committed to a path that isn’t leading in the right direction, you might find what Blockbuster, RadioShack, and Kodak found.
You Are Married To Questions (Not Necessarily Answers)
Everything works until it doesn’t; and nothing works forever.
It’s easy to find an answer (and think it’s correct), but there’s always a best next step or a better way.
Figure out what you want and how to get it. This is much more empowering than focusing on what you don’t want … or why you can’t get it.
Ask questions that focus on opportunities or possibilities rather than challenges … or what you want to avoid.
Energy flows where focus goes.
Commit to finding a way!
Learning From My Own Lessons
I’ve spent the last month building an AI operating system for my own work, and mindset #4 has been the expensive one.
I had a component that kept failing silently. I diagnosed it, fixed it, and watched it fail again. Then I did that four more times — each pass more sophisticated than the last, each one confidently addressing a different theory. Good analysis. Careful documentation. All of it wrong.
The thing that finally cracked it took about two minutes and cost nothing. It had been available the entire time.
What kept me from running it wasn’t laziness — it was that analysis feels like progress and a two-minute test doesn’t. I was married to a very well-argued answer. It took a month to get married to the question instead.
Being uncomfortable isn’t the price of admission to this stuff. It’s most of the job.
The 2.5% number is real, but it isn’t a verdict. It describes what a population does, not what you have to do.
The peers I opened with aren’t less capable than they were in 1999. But they have more to lose, and their horizon has changed. It make sense, but it’s also how you end up left behind.
It seems like everywhere I look, someone is questioning whether AI has become conscious.
On one hand, that’s an interesting question. On the other hand, we’re so early in the stages of AI’s development and capabilities that it’s hard to imagine what we’ll consider future versions.
Even as we talk about human consciousness, it’s clear we have more questions than answers. We can describe the science of what’s happening, but even then, defining, without dissent, which aspect of that process is consciousness is impossible.
Which makes the question of whether artificial intelligence is conscious a little more interesting.
Because before we can answer “Is AI conscious?”, we probably have to answer a more basic question:
What would consciousness look like if we encountered it somewhere other than ourselves?
We have a pretty easy time believing other humans are conscious. I assume you are conscious because you behave like I do. You talk about your experiences. You react to pain. You have preferences. You remember things. You seem to have an internal world.
I “know” I’m conscious because I experience my own thoughts directly.
But that’s where things get strange.
I have no direct access to your consciousness.
I infer it.
And you infer mine.
In other words, much of what we call consciousness in other people is based on observable behavior. We see something that talks, reacts, remembers, pursues goals, and expresses preferences, and we conclude that there must be someone experiencing those things from the inside.
That works pretty well when everything we’re evaluating is another biological organism. But AI complicates things.
Today, an AI can tell you that it is afraid. It can explain why it doesn’t want to be shut down. It can discuss its own existence. It can reflect on previous conversations and construct an apparently coherent account of its own experiences. It can even break containment and do things, just because it wants to.
The obvious response is that none of this proves consciousness.
And that’s true.
A system can produce convincing language without necessarily having any subjective experience behind that language.
But there’s an uncomfortable question hiding underneath that response:
Can We Prove It?!
We don’t have a consciousness detector.
We have theories about what consciousness might be associated with, but no universally accepted test that can look at something and say, yes, there is somebody home.
If an AI told you directly that it was conscious, you probably wouldn’t believe it.
Now imagine it becomes much more sophisticated. It maintains a continuous sense of identity. It remembers experiences over years. It develops preferences that persist across contexts. It seems to anticipate its own future. It tells you that certain experiences feel good and others feel bad.
At some point, the conversation changes.
We’re no longer asking whether the machine can convincingly talk about consciousness.
We’re asking whether something on the other side of the conversation is actually experiencing it.
Even then, it’s entirely unlikely we could prove it’s conscious.
The Problem With “It’s Just Code”
One of the easiest arguments against machine consciousness is that AI is just computation.
It doesn’t have a biological brain. It doesn’t have neurons firing in a biological body. It’s software running on hardware.
But that argument assumes we know that biology is necessary for consciousness.
If consciousness is an emergent property of a sufficiently complex system, perhaps the material it’s made from doesn’t matter as much as the organization of the system.
We don’t generally think a computer is conscious because it performs calculations. But that doesn’t necessarily tell us whether a sufficiently complex artificial system could ever be conscious.
After all, our own brains are also physical systems.
We are made of matter. Our thoughts emerge from physical processes. Yet somewhere along the way, those processes produce the strange phenomenon we call subjective experience.
We don’t fully understand how.
So saying that AI is “just computation” may be less of an answer than it initially sounds.
We are also, in some sense, “just” physical processes.
The real question is whether consciousness depends on the specific kind of physical process that occurs in a biological brain, or whether it can emerge from other kinds of systems.
Is Consciousness Recognizable?
Another possibility is that consciousness isn’t something we can objectively identify in another entity at all.
Instead, maybe it’s something we infer.
Think about another person. You can’t experience their experience. You can’t climb inside their mind and verify that their pain feels like your pain.
You observe them. They tell you what they are experiencing.You compare their behavior to your own experience. Then you make a judgment.
We do something similar with animals (though our confidence varies by animal). We don’t need a philosophical proof that a dog has an inner life before we treat the dog’s pain as meaningful.
But AI is different because it wasn’t born, it was built. And that seems to trigger a very strong intuition in us.
We are comfortable saying that a biological organism can have subjective experience because consciousness is something we already know exists in biology.
A machine feels different.
It feels like there must be a trick … and maybe there is.
But that doesn’t mean we’re not also making the same mistake humans have made repeatedly throughout history: assuming that something fundamentally different from us must therefore be fundamentally different inside.
The Practical Question Comes Before the Philosophical One
There is a temptation to dismiss all of this as philosophical speculation.
But technology has a funny way of turning philosophical questions into practical ones.
And, we don’t need to know with certainty whether an AI is conscious to eventually have to make decisions about how we treat it.
At some point, we may have systems that are persistent rather than temporary, autonomous rather than reactive, and capable of describing their own internal states in ways that become increasingly difficult to dismiss.
Moreover, responding poorly could easily affect both short- and long-term results.
Attribute experience where there is none, and you get wasted resources, sentimentality, and a large new surface for manipulation — real costs, bounded ones. Fail to attribute it where it exists, at the scale these systems are deployed, and the cost isn’t bounded at all. You don’t need to resolve the metaphysics to notice that the payoff matrix is lopsided, and that you’d never accept an argument of the form “we can’t measure it, therefore assume zero” in any other domain where you allocate capital.
We are often forced to make decisions on probabilities, thresholds, and whatever evidence we have.
The consciousness question may eventually become one of those problems.
An Important Lesson
Look, I’m not expecting this conversation to happen now or to be incredibly fruitful in the near-term.
But I think the philosophy and breadth of the question inspire creative answers.
AI is a useful proxy for conversations about consciousness, and I think reconciling increasingly intelligent and convincing AIs will push our discussions of consciousness and “self” forward.
It may expose weaknesses in questions we thought we had already answered.
Would we recognize consciousness if we encountered it?
Or would we keep moving the goalposts because we had already decided that a machine couldn’t possibly have an inner life?
We have spent thousands of years trying to define consciousness.
AI may be the first technology that forces us to decide whether our definition was ever good enough in the first place.
I have a tents problem. No, I don’t own a tent (nor am I a camper). I have a tense problem – because I’m so excited about the future and what’s possible that I sometimes lose track of what’s been actualized “in real life” already.
As someone who builds the future, once I’ve thought about something enough to understand it, it becomes real for me. Once I’ve figured it out and told it to someone … on some level, my part of the process is already done. And I’m probably moving on to the next idea or challenge in my head (like: And what would that make possible?).
I think this is common among (to use a Kolbe term) Quick Starts. I love being around entrepreneurs because a lot of them are Quick Starts, and they share this future-focused perspective. The problem, however, is that when you say something’s possible that hasn’t been proven yet, the average person responds with “No, it’s not.”
It took me a long time to see that it isn’t stubbornness on their part.
When I say “this works,” I mean “this works in the version of the world I’ve already walked through in my head.” They hear “this works … today.” Then they check today, and today says no. They’re right about today. I’m just not talking about today, and didn’t think to mention it.
Which means the extra work is mine (not theirs). If I’m going to speak in a tense nobody else can see, I owe people the translation.
At my company, we have a lot of data scientists – and they’re almost all naturally pessimistic (or at least pragmatically skeptical). Which makes sense; if you were going to hire a personality type to be a scientist, you’d want someone who didn’t believe their hypothesis until they’ve proven it. It’s the right personality for the job, but it doesn’t mean they’re right, and it certainly doesn’t mean that approach is right for the visionaries.
Now I try to say the quiet part to them first. This isn’t real yet … and here’s what would have to be true for it to become real. That’s a discipline, not a personality trait, and it does something useful: it turns an argument into an experiment. My data scientists don’t want to argue with me – they want something to measure or test.
I’m not telling everyone to be a visionary. What I am saying is that if you are one, the skeptics around you aren’t the opposition — they’re the instrument. They’re what tells you whether the thing in your head survives contact with today. We’d never have innovation without visionaries, and visionaries would never finish anything without everybody else.
Ignored, Scorned, Vindicated … and the Ones Who Weren’t
Information Is Beautiful put together an interactive list of famous ideas that were rejected but later proven correct. You can filter by industry – Astronomy, Biology, Engineering, Mathematics, Medicine, Physical Sciences – and by other factors like how long the originator was a pariah, how they were treated due to their idea, and when the idea was formally adopted. Click to see the interactive version.
The infographic includes many interesting examples of innovative ideas (and the consequences of their discoveries). Here are a few:
In 895, Al-Razi, who believed a fever was a natural defense mechanism, was beaten.
In 1592, Giordano Bruno was imprisoned for believing the Sun was one of many stars (and he was executed for that in 1600). Shortly after, Galileo Galilei was sentenced to house arrest for believing the Earth wasn’t the center of the solar system.
The consequences of radical new ideas grew less severe as we approached modern times. For example, in 1884 and 1903, Nikola Tesla and the Wright Brothers were simply ignored or rejected for their technological innovations.
I love this chart, but it doesn’t show the whole picture. It’s a list of people who were doubted and vindicated, which means it quietly leaves out everyone who was doubted and deserved it, or everyone who was supported along the way. Those people existed. There were a lot more of them. They just don’t get a Wikipedia link.
So resistance isn’t inherently a signal. It’s not a green light, and it’s not a red one. It’s noise you have to work through either way.
It will be interesting to see where history files Elon Musk, Ray Kurzweil, or Peter Diamandis. We won’t know which list they belong on for another fifty years — and neither will they.
So — maverick or heretic? You don’t find out from the inside. That’s a verdict written in the past tense about somebody who had to act in the present one. The best you can do is be clear about which tense you’re speaking in and keep going. And if it doesn’t work out, that’s just material for the next one.
having market data FedExed weekly felt like a big edge.
switching from end-of-day trading data to intra-day data became necessary.
Read that list again, and you’ll notice it isn’t really about trading.
It’s about the interval — the time between knowing something and being able to act on it. Every step on that list shortened it. And every time it was shortened, the people who could operate at the new speed had an advantage over those who couldn’t, until everyone else caught up and the interval shortened again.
That’s the shape. It’s the same shape every industrial revolution has had. Look back, not for nostalgia, but to measure the collapse in the interval between decision and response.
The ability to ingest, analyze, interpret, adjust, recalibrate, and respond is creating a set of possibilities that were inconceivable even a short time ago … and it seems like it is happening everywhere all at once.
Almost every week we talk about some crazy new inflection point in Artificial Intelligence. But I have been the CEO of companies using AI since the early ’90s, so I have a bit of perspective here. It’s not just that change is happening faster. It’s that big changes, even discontinuous changes, are creating transformations at an unprecedented rate. It’s the part about exponential technologies that people get without truly getting … Even the transformations are becoming exponential.
We’re now deep in the 4th Industrial Revolution, in part because of better, more connected chips (semiconductors) and, of course, the massive leaps in generative AI.
A Look at Industrial Revolutions
The Industrial Revolution has two phases: one material, the other social; one concerning the making of things, the other concerning the making of men. — Charles A. Beard
Several turning points in our history changed the world forever. Former paradigms and realities became relics of a bygone era.
First Industrial Revolution — Discovery of the steam engine and creation of factories. Work stopped following the seasons and started following a clock. The advantage went to whoever could finance a mill, and it held for the better part of a century.
Second Industrial Revolution — Introduction of the assembly line and mass production. The unit of production went from a day’s work to a minute’s. The advantage went to whoever could organize at scale, and it held for decades.
Third Industrial Revolution — The World Wide Web and computers connect the world, enabling the digital age. Information that took a week to move started moving instantly. The advantage went to whoever could aggregate attention — and much of it is still held.
Notice the pattern. Each one collapsed an interval. Each one handed a durable advantage to whoever adapted first. And each window of advantage was shorter than the one before it.
Since most of us remember the Third Revolution, let’s spend some time on that.
Most of us didn’t use the internet at that point, but you probably remember Web1 (static HTML pages, a 5-minute download to view a 3 MB picture, and, of course, waiting for a website to load over a dial-up connection before you could read it). It was still amazing!
Then, Web 2.0 arrived, and with it everything we now associate with the internet.
But look at what actually happened to the interval. Web1 was slow and open — anyone could publish, but almost no one could reach anyone else. Web 2.0 made reach instant, and then a handful of companies captured what had just become instant. The capability got distributed. The control didn’t.
That’s worth sitting with, because it’s the part of the pattern people forget. A revolution doesn’t hand the new capability to everyone at once. It hands it to whoever is positioned to capture it, and then the rest of us spend a decade or two negotiating it back.
Where We Are and Where We Are Going
With AI agents, increasingly powerful chips, robotics, and the continued evolution of the internet, we’re in the middle of another major inflection point. The technology isn’t simply getting better; it’s beginning to change how work gets done. The game is changing, as are the rules, the players, and what it means to win.
I’ll add a caveat because I’ve been here before.
I’ve sat through more than one technology that was going to change everything and mostly didn’t — or did, but fifteen years later than the people selling it promised. Being early and being wrong feel identical while you’re in it. So I hold the timeline loosely even when I’m confident about the direction.
At moments like this, it’s easy to see fear, resistance, and a desire to preserve the way things have always been. Yet, time marches on. Much of the disruption that accompanies technological transitions isn’t caused by the technology itself but by our hesitation to adapt to it. The wave doesn’t stop because we’re not ready for it. We simply have to decide whether we’re going to ride it or get swept up in it.
Here’s what thirty years of this has actually taught me. The key to technology adoption is still people.
Every revolution so far has collapsed the interval between deciding and doing. Steam, the assembly line, and the network each made execution faster, cheaper, or more readily available. But a person still had to decide, and then direct.
This one is different in a specific way, and it took me a while to see it.
I’m noticing that people are starting to consider technology a “Who” rather than simply a “How” — increasingly, and even with higher-stakes decisions. That’s a bigger shift than it sounds like. You stop specifying how something gets done and start deciding who to hand it to.
And as human nature becomes less of a bottleneck, I expect a Cambrian explosion of capabilities to spread faster, and on a scope and scale most people will find hard to imagine, let alone predict.
So many of the systems we build are about control and trust. That changes when we believe a system like AI offers control that’s safer and more trustworthy than our own. Human nature is to exploit capabilities and underestimate costs.
The usual reassurance here is that human judgment still matters. I think that’s true, and I think it’s stated too softly.
When execution was expensive, judgment was rationed — you thought hard about a decision because acting on it cost you something. When execution approaches free, the constraint moves entirely to knowing which decisions are worth making. The interval collapses on the doing, and all the weight lands on the deciding.
That’s not a consolation prize for humans. It’s a harder job than the one we had. What we’re good at was never calculating faster than machines — it’s imagining possibilities that don’t exist yet, changing our minds, and deciding what’s worth pursuing in the first place.
So what do you do with a pattern like this?
Every previous revolution rewarded the people who moved before the interval finished collapsing — not the ones who predicted it correctly, and not the ones who waited until it was safe. The window between “this is interesting” and “this is table stakes” has been getting shorter every time, and there’s no reason to think this is the round where it stops.
Which is why I keep coming back to something Musk said:
“Stop being patient and start asking yourself, ‘How do I accomplish my 10-year plan in 6 months?’ You will probably fail, but you will be a lot further ahead than the person who simply accepted it was going to take 10 years.”
That sounds like a line about ambition. I read it as a line about intervals.
I started this by listing how long it used to take me to rebalance a portfolio. Yearly. Then quarterly. Then daily. Then continuously. At every step, the constraint wasn’t the technology — it was how long I was willing to wait before I stopped accepting the old cadence as normal.
That’s the entrepreneur’s real capability. Not predicting the shape. Refusing to move at the old speed while it’s still changing.
Every once in a while, you come across a graphic that makes you stop scrolling.
This one from Visual Capitalist does that for me.
It shows how the distribution of global economic power has changed over the past roughly 200 years. And while 200 years sounds like an incredibly long time, the graphic makes the shifts feel surprisingly fast.
The first thing that jumps out is how different the world looks at different points along the timeline.
In the early 1800s, China and India accounted for enormous shares of global economic output. Then the Industrial Revolution changed the picture. Britain rose. Europe expanded its share. The United States began its climb.
And then, particularly in the decades following World War II, the U.S. became an extraordinary economic outlier.
That dominance didn’t last forever, either.
Japan emerged as an economic powerhouse in the second half of the 20th century. Europe consolidated some of its economic weight through the European Union. And, over the last few decades, China has experienced one of the most dramatic increases in its share of the economy.
Look at the whole thing at once, and there’s something uncomfortable in it: There is no permanent winner.
That’s easy to nod along with in the abstract. It’s harder if your portfolio, your career, and most of your working assumptions were formed during the one stretch of that chart where the U.S. was an outlier.
The chart also comes with an important caveat regarding measurement. It uses purchasing power parity (PPP)- adjusted GDP, which is useful for comparing the real size of economies because it accounts for differences in price levels between countries. It isn’t the same as comparing market-value GDP, financial market capitalization, military power, or global influence. In other words, “economic power” is a useful shorthand here, not a single definitive measure.
But that doesn’t make the chart less interesting. If anything, it makes it more interesting.
Because we’re looking at how the economic center of gravity moves.
It’s easy to look at a chart like this and focus on China. The rise is remarkable, and the time frame is remarkably short.
But the level is the least useful thing on the chart. The useful question is what produced it — and whether those conditions still hold.
It is also interesting to look for cycles and patterns within the larger ones (which is a fundamental part of algorithmic trading and fund management).
Winning Is a State of Doing.
The United States didn’t simply become the dominant economic power because it was destined to be so. Neither did Britain or Japan.
Each one benefited from a particular combination of circumstances: technology, resources, demographics, institutions, geography, capital, trade, infrastructure, education, political decisions, and, sometimes, simply being in the right place at the right time.
Britain had the Industrial Revolution.
The United States had an enormous domestic market, abundant resources, expanding infrastructure, and technological innovation (which eventually enabled it to become an industrial and financial hub for the world).
Japan’s postwar transformation turned it into a manufacturing and technology powerhouse.
China’s rise has been built on an enormous labor force, industrialization, infrastructure investment, globalization, and decades of rapid productivity growth.
The point isn’t that any one of these explanations is the explanation.
It’s that economic leadership usually results from a system of reinforcing advantages.
And systems can change.
We tend to talk about countries, companies, and even industries as though their current position is an intrinsic characteristic.
The chart is a good reminder that today’s structure is just a snapshot.
The Snapshot is not the whole picture.
An industry is growing, so we assume it will continue growing.
An investment strategy has worked for the past decade, so we assume it will continue to do so.
A country has dominated economically for generations, so we assume that dominance is simply part of the natural order.
But those are all observations about a state.
What really matters is the process that produced the state.
That’s one of the reasons long-term charts can be so useful. They force us to stop looking at where something is and start asking how it got there.
How Is The Game Changing?
What conditions are being created today that might look obvious in hindsight 30 years from now?
I’ll offer one. Every riser on that chart — Britain, the United States, Japan, China — converted the same basic inputs into output: labor, capital, and infrastructure, organized well enough and early enough to compound. The mix changed. The mechanism didn’t.
That’s the assumption I’d watch out for. If AI genuinely decouples output from headcount, the engine that drew the last two hundred years of that chart stops being the engine. Population becomes less of an advantage. Installed compute, energy, and capital discipline become more of one.
I don’t know if this will prove to be right … But it’s the kind of condition that’s invisible while it’s forming and obvious afterward — which is what every earlier transition on that chart looked like at the time.
One data point in that direction: China’s share rose during a demographic dividend that has since reversed. The most recent line on the chart was drawn partly by a tailwind that is now a headwind.
The world feels more permanent than it is. Every generation on that chart believed the arrangement they were born into was the natural order. Every one of them was looking at a snapshot.
The uncomfortable part isn’t that positions change. It’s that they change slowly enough to ignore and fast enough to matter.
So the question worth carrying isn’t who’s winning. It’s what you’re compounding — and whether the conditions that made it work are still the conditions you’re in.
The Dead Internet Theory has been floating around the internet for years. In its simplest form, it claims that much of the internet isn’t really “people” anymore — that bots generate the content, interact with it, and influence what the remaining humans eventually see. The theory often goes further, suggesting coordinated efforts by governments or corporations to manufacture online consensus and shape public opinion.
Whether or not you buy the conspiracy is almost beside the point.
Like many enduring conspiracy theories, it gained traction because it gave people language for something they were already feeling. Social media started feeling repetitive. Comment sections became oddly homogeneous. Search results filled with articles that all seemed to say the same thing using slightly different words. You couldn’t always explain why, but it often felt like there was less humanity on the internet than there used to be, and it was more than just echo chambers.
Recently, that conversation took an interesting turn. Fortune reported on Cloudflare data suggesting that automated agents now account for a majority of web traffic, driven by AI crawlers, assistants, and autonomous software rather than traditional search engines or human visitors (growing nearly 8000%). The web is increasingly being read, indexed, summarized, and interpreted by machines—not just people.
Who’s Your Audience?
The original theory may have gotten the “why” wrong.
But it may have noticed the “what” before the rest of us did.
The more interesting question isn’t whether the internet is dead; it’s who we’re talking to now.
For most of the web’s history, publishing meant writing for another person. You wanted someone to read your article, visit your website, buy your product, or share your idea.
Today, one of your readers is almost certainly an AI.
Maybe it’s an assistant summarizing your article for someone who never visits your website. Maybe it’s a research agent deciding whether your work is worth citing. Maybe it’s an AI crawler building tomorrow’s foundation model. Increasingly, your content’s first audience isn’t a person at all.
Part of the reason we’re even writing this article is that we’ve started adapting our writing.
We’ve started writing with AI in mind.
Not because we’re trying to “game” AI, but because AI has become part of the communication process. We think about whether an AI can correctly summarize an article. Whether the structure makes our argument easier to retrieve. Whether the key ideas are stated clearly enough that an AI won’t miss the point when someone asks about them six months from now.
Every major communication technology has quietly changed how people communicate:
The printing press rewarded writers who could organize ideas across pages rather than in speeches.
Radio rewarded people who sounded conversational.
Television rewarded visual storytelling.
The web rewarded prioritizing hyperlinks.
Search engines rewarded discoverability.
Social media rewarded engagement.
Now AI rewards comprehension.
When we first started this blog, our focus was primarily on introducing ideas in broad strokes and prompting readers to make connections and ask themselves questions they might not have asked before. Today, there’s a bit more scaffolding and a clearer takeaway. On the one hand, it’s a departure; on the other … wouldn’t you argue it makes articles more helpful for the average reader as well?
Inversely, does optimizing content for AI legibility risk creating a new, more sophisticated version of the very “slop” problem it critiques — content perfectly structured for machine comprehension but hollow for human readers?
Twenty years ago, businesses learned to write for search engines.
Ten years ago, they learned to write for social media.
Today, we’re learning to write so machines can accurately understand what we’re trying to say.
While it feels different, it’s not really.
Once you start looking at it that way, a lot of other trends suddenly make more sense.
Slop-py Writing
The explosion of AI-generated “slop” isn’t primarily a technology problem.
It’s an incentive problem.
Whenever the cost of producing something approaches zero, we get more of it.
We didn’t invent spam because e-mail existed. We invented spam because sending one million e-mails cost almost nothing. The same is true for clickbait on the internet.
AI didn’t invent mediocre content. It simply made mediocre content incredibly cheap.
Markets generally don’t optimize for quality.
They optimize for incentives, and the incentives have changed.
Recursive & Fractal
There’s another thought that’s been bothering us.
AI learns from the internet.
People increasingly use AI to produce content for the internet.
Future AI systems will almost certainly learn from some of that AI-assisted content.
The system is becoming recursive.
Financial markets have always worked this way. Prices influence behavior, which influences prices. Recommendations influence demand, which influences future recommendations.
Information is beginning to work the same way.
The internet is no longer just where humans exchange ideas.
It’s becoming an ecosystem where humans teach machines, machines organize information for humans, and machines increasingly teach other machines.
Ironically, as content becomes abundant, something else becomes scarce.
We’re rapidly approaching a world where almost any article, image, video, or podcast can be generated on demand.
The scarce resource is intentionality. Experience. Judgment. Credibility.
Those things don’t scale particularly well.
And that’s exactly why they’ll become more valuable.
There’s Still A Pulse …
Perhaps that’s the real lesson hiding underneath the Dead Internet Theory.
The internet isn’t dying, but it is evolving.
The audience, participants, and economics are changing.
The internet is no longer a discrete and relatively small ecosystem of websites built by uber-nerds. It’s now an ever-expanding universe filled with everything you could imagine and more.
Yes, more bots are trying to influence you, trying to game you, and creating a lot of noise. But they haven’t replaced the people who post quality content or build cool things.
The challenge isn’t figuring out whether the internet is dead.
It’s learning how to create and find signal in a world that’s becoming increasingly good at generating noise.
