The Internet Is Splitting Into Two Types of People

It has nothing to do with how smart you are. It has everything to do with what you do with the hours.

I’ve been thinking about a thought experiment that I can’t stop coming back to.

Two people wake up tomorrow and spend the same eight hours online. Same internet. Same tools. Same access to the same information, the same AI models, the same platforms, the same opportunities that have never existed at this scale in the history of human civilization. By evening, they are further apart than they were when they woke up — and the gap will be larger again tomorrow.

Photo by Robert Anasch on Unsplash

The first person leaves with a new skill they actually practiced, a piece of writing they published, a workflow they automated, a conversation that opened a door they didn’t know existed. They didn’t finish everything. But something in the world is different because of those eight hours.

The second person closes their laptop carrying a residue of outrage from a political argument they didn’t start, exhaustion from content they can’t quite remember, a vague awareness that they’ve been sitting for hours with nothing to show for it, and a low background anxiety that everyone else is moving faster.

Same internet. Completely different trajectories.

This isn’t a story about discipline or willpower — those are the wrong frames entirely. It’s about something more fundamental: what the internet has quietly become, and what it’s doing differently to different kinds of people.

The internet used to connect us. Now it’s sorting us.

I. What the internet actually is now

There’s a version of internet history that most people carry around in their heads, and it goes like this: the internet democratised information, which democratised opportunity, which is generally a good thing for humanity.

That story was true for about twenty years. It stopped being true around 2010.

The original internet was genuinely about access. If you lived in a small town and wanted to learn about medieval European history, or how to write a business plan, or what a derivative was, the internet gave you access to that information in a way the library couldn’t. Information was the bottleneck. Finding it was the skill. The person who could locate, synthesise, and apply information faster than others had a real advantage.

That advantage is gone.

Information is no longer scarce. There is more high-quality, freely available information on any subject you can name than any human being could consume in a hundred lifetimes. Every concept explained from first principles. Every skill documented with tutorials. Every industry analysed by practitioners. And on top of all of that, AI systems now exist that can retrieve, synthesise, and explain almost anything on demand, in plain language, in seconds.

The bottleneck moved.

It was once: can you find the information?

Now it is: what do you do with it?

Information is no longer power. Applied intelligence is.

The person who merely finds information — who reads the article, watches the tutorial, saves the thread — gains almost nothing. The person who applies it, tests it, builds something with it, publishes the result, and iterates on the feedback — gains compounding advantage every time they do it.

This is not a minor shift in how the internet works. It is a complete inversion of the original value proposition. And most people haven’t updated their mental model to reflect it.

II. The two groups that are emerging

Every technology in history has divided its users into two rough categories.

When electricity arrived, some people used it to build factories and industrialise production at scale. Others used it to illuminate the same activities they were already doing. Both had access. The outcomes were not equivalent.

When printing arrived, some people used it to publish ideas that changed how millions of people thought about religion, politics, and science. Others used it to replicate handwritten manuscripts faster. Both had access. The outcomes were not equivalent.

The internet is doing the same thing now, but faster and at greater scale than any previous technology. It is creating two groups that have access to the same tools and are using them in fundamentally different ways — and the gap between the groups is widening faster than most people recognise.

The first group — builders

Builders treat the internet as a production environment. Something goes in, something comes out.

They open a browser with an intention. They use AI to think faster, not to avoid thinking. They publish what they learn because they know the act of explaining forces clarity. They automate the repetitive parts of their work so they can spend more time on the parts that require judgment. They build things — products, content, systems, tools, skills — that exist and compound after the session ends.

The builder’s day looks something like this. An hour learning something new, not by watching someone else learn it but by trying it themselves and hitting the edges of what they don’t understand. Another hour creating — writing, building, shipping, publishing, recording — something that didn’t exist before. They use AI as a collaborator, not a replacement. They end the day with something to show for it.

What compounds for builders: skills, reputation, relationships, income streams, systems, leverage. Each thing they build makes the next thing easier to build. This is the flywheel that most people can see operating in the lives of people they admire from a distance, but can’t quite trace back to its source.

The second group — passive consumers

Passive consumers treat the internet as a delivery mechanism for sensation.

They open a browser without a clear intention and let the algorithm choose what happens next. They consume enormous volumes of information without processing it into understanding. They watch people build things without building anything themselves. They collect ideas without applying them. They mistake familiarity with knowledge — the feeling of having read about something is genuinely similar to the feeling of having understood it, and the brain doesn’t always distinguish between the two.

The passive consumer’s day looks something like this. They open their phone before they’ve had a thought of their own. They spend hours moving between platforms, each one optimised by engineers to make leaving feel like the wrong choice. By the end of the day, they’ve absorbed hundreds of fragments of information from dozens of sources, and the net result is approximately nothing — no new skill, no new output, no new asset, no new relationship, just the cumulative fatigue of sustained low-level stimulation.

The passive consumer learns about building. The builder builds.

This is not a moral judgment. Passive consumption is seductive precisely because it delivers genuine short-term pleasure. The content is often interesting. The outrage is genuinely stimulating. The entertainment is genuinely entertaining. The problem isn’t that passive consumption feels bad. The problem is that it doesn’t compound — and everything about modern economic life rewards things that compound.

III. Why AI doesn’t help everyone equally

Here is the thing about AI that most AI coverage gets wrong.

The coverage treats AI as a great equaliser. It says: now everyone has access to the same intelligence, the same research assistant, the same coding partner, the same writing collaborator. The playing field is being levelled.

This is precisely backwards.

AI is not a great equaliser. It is a great amplifier. And amplifiers don’t level fields — they make existing differences larger.

Give a curious, active person access to an AI assistant and what happens? They learn faster. They prototype ideas in hours instead of weeks. They ask better questions and get better answers. They use the tool to do things they couldn’t do before — write code they couldn’t write, design systems they couldn’t design, research topics they couldn’t have researched without weeks of work. Their existing habits of building and experimenting are made more powerful.

Give a passive consumer access to the same AI assistant and what happens? They use it to generate content they don’t engage with deeply, to get answers they don’t apply, to have conversations that produce no output. The AI becomes another source of content — a more personalised, more responsive one, but fundamentally another thing to consume rather than a tool to build with.

The variable isn’t the AI. It’s the habit of mind the person brings to the tool.

Claude, ChatGPT, Gemini, Perplexity, coding assistants, research tools — these are genuinely extraordinary capabilities, available to almost everyone for close to free. They are the most powerful tools for expanding human capability that have ever existed. And they are making the gap between builders and passive consumers larger every month, because builders use them to build and passive consumers use them to consume, and the compounding runs in both directions.

The tool becomes what you already are, but more so.

IV. What actually creates value now

For most of the industrial era, the valuable inputs were obvious. Physical labour. Memorised knowledge. Credentials that signalled competence. Time spent in a building doing a defined task.

The economy rewarded those things because they were scarce and because they were measurable. You could count the hours worked. You could verify the degree. You could observe the physical output.

AI is making most of those things abundant. Physical presence is increasingly unnecessary for knowledge work. Memorised information is worth less than the ability to ask good questions of a system that has access to all of it. Credentials still open doors, but they no longer guarantee what’s behind the door.

What remains scarce — genuinely, increasingly scarce — is something harder to measure and harder to copy.

Attention — not the ability to pay attention for long periods, but the ability to direct attention deliberately toward the things that compound. The most valuable skill in an environment designed to steal your attention is knowing where to point it.

Creativity — not artistic creativity necessarily, but the ability to combine existing things in ways that produce something new. This is something AI actively cannot do well. It can recombine. It cannot originate. The person who brings an original perspective to a problem — who sees something nobody else has articulated yet — has an advantage that grows as AI handles more of the routine work.

Judgment — knowing which question is worth answering, which analysis matters, which direction to move. This is the layer above tools. AI can execute. AI cannot judge. The ability to make good calls under uncertainty, with incomplete information and real stakes, is worth more than it has ever been.

Adaptability — the rate at which you update your model of the world when new information arrives. People who are attached to how things worked are losing ground faster than any previous generation has had to experience. The people gaining ground fastest are the ones who find the process of change interesting rather than threatening.

Execution — the ability to start, continue, and finish things. Ideas are everywhere. Applied ideas are rare. The person who ships imperfect things consistently beats the person who waits for the perfect version indefinitely.

Degrees opened doors. Portfolios prove what’s behind them.

The new career path doesn’t follow a credential → application → job → promotion sequence. It follows a build → publish → iterate → opportunity sequence. The building is public. The proof is in the work itself. And the work accumulates in ways that credential sequences do not.

V. The builder mindset — what it looks like in practice

The builder mindset is not a personality type. It is a set of habits applied consistently over time.

Some of the most productive people using the internet as a building environment are introverts who never imagined themselves as public creators. Some are people who failed academically and found that building things was the first time their particular way of thinking was valued. Some stumbled into it accidentally and gradually realised that the feeling of making something was different from the feeling of consuming something — in ways that made them want more of the making.

The common thread isn’t personality. It’s practice.

Here’s what the builder mindset looks like for different kinds of people:

For creators and analysts: instead of reading twenty articles about a topic, read three and write one. The writing forces you to identify what you actually understand versus what you merely recognise. Publish it, even imperfectly, because the feedback loop between publication and reality is the only loop that actually sharpens your thinking over time. The analyst who writes about what they observe builds a public record of their thinking that compounds. The one who only reads builds nothing external.

For developers: instead of watching tutorials end-to-end before touching a keyboard, build the broken version first. Find the specific thing you don’t understand. Ask the AI about that specific thing. Apply the answer to your specific broken version. This is four times faster than the tutorial-first approach and produces understanding rather than familiarity. Side projects are how developers build judgment that their job alone cannot give them.

For entrepreneurs: instead of collecting frameworks about how to build a business, build the smallest possible version of one thing and put it in front of one person. The feedback from one real person interacting with one real thing is worth more than a hundred frameworks consumed in theory. The internet gives you distribution before you have a product — which means you can build an audience for something that doesn’t exist yet, and then build for the audience you already have.

For students: instead of using AI to get answers, use it to generate questions. Ask it to challenge the thing you think you understand. Ask it what the strongest argument against your current view is. Ask it to help you design a project that would require you to apply what you’re learning rather than just absorb it. The student who uses AI to avoid thinking will know less in four years than the one who uses AI to think more deeply and quickly.

VI. The framework — consume, think, build, share, improve

Most internet behavior follows a one-step loop: consume, repeat. The builder’s loop has more steps, which is exactly why most people don’t follow it. The friction is in the middle.

Here is the full loop:

Consume — deliberately, not passively. Choose what you put in your head the same way you choose what you put in your body. Not because everything should be optimised for productivity, but because the inputs shape what outputs are possible. Consume things that generate questions in you, not just things that generate reactions.

Think — before you move to the next piece of content, pause. What does this change about how you understand something? What question does this raise that you don’t have an answer to? What would you do differently if this were true? Most people skip this step entirely. The pause is where understanding begins.

Build — make something. This doesn’t have to be large. A paragraph. A prototype. An email to someone. A small piece of code. A document that organises your current thinking. It can be rough. It needs to exist. The act of making forces a kind of clarity that consuming never produces. You discover what you don’t know by trying to apply what you thought you knew.

Share — put the thing in front of at least one other person. This step is the one most people skip because it feels risky. Sharing means being evaluated. It means being wrong in public. It also means getting the only feedback loop that actually works — the loop between what you made and how it lands in the real world. Everything before sharing is practice. Sharing is where learning becomes real.

Improve — use what came back from sharing to make the next version better. This is the compounding step. Each cycle of the loop produces something slightly better than the last, slightly faster than the last. The compound rate looks small in the first few cycles and enormous by the hundredth.

Repeat — not occasionally, not when motivated. Consistently. The loop needs to become a habit, which means it needs to run when motivation is absent, which means the steps need to be small enough that the barrier to starting is low even on bad days.

Every day you do the loop is a day you move forward. Every day you skip it is a day someone else who doesn’t is moving forward instead.

The gap between people running this loop and people not running it doesn’t look like much after a week. It looks significant after six months. It looks insurmountable after three years. This is how people who seemed ordinary end up somewhere that seems extraordinary, and how the people who watched them from a distance have a hard time explaining what happened.

The honest close

The internet is not neutral. It never was, but the stakes are higher now than they used to be, and the feedback loops are faster.

There are two versions of you that the internet is capable of producing.

One version spends hours each day being fed content chosen by systems whose entire purpose is to maximise the time you spend being fed content. This version ends most days slightly more anxious, slightly more reactive, slightly more dependent on the next stimulus. Over years, this version becomes easier to predict, easier to influence, and paradoxically — in a world that theoretically offers unlimited opportunity — more limited in actual possibility.

The other version treats the internet as a workshop. Something goes in. Something comes out. This version builds things that outlast the session, develops judgment that compounds, and creates a reputation and a portfolio of work that acts as leverage in ways that passive consumption never can.

The difference between these two versions of you isn’t intelligence. It isn’t work ethic in the conventional sense. It isn’t even time — both versions spend roughly the same hours online.

It’s what those hours are for.

You don’t need to eliminate consumption. You don’t need to become a content machine. You need to tip the balance enough that you end most days with something that didn’t exist in the morning — a piece of writing, a small project, a question answered by doing rather than reading, a thing in the world that carries your name and your thinking and compounds quietly while you sleep.

The internet won’t decide your future.

Your habits will.

Every click is a vote for the person you’re becoming.

I write about how AI, technology, and behavior shape careers and lives — in plain language. Follow for more.