The Rabbit Hole, Part 3: The Internet Poured Rocket Fuel on It
Expertise used to be gated by geography. Online enthusiast communities collapsed the path into any hobby from years to one evening. Here is what it cost.
Type “burr geometry” into a search box tonight. By midnight you will know the difference between conical and flat, why one camp swears flat burrs give more clarity in the cup, which grinder the people who actually care consider the honest value pick, and exactly where on the internet those people argue about it.
You will not own a grinder. You will not have made a single cup of coffee. You will have the entire vocabulary.
That gap between knowing the language and having the skill is new. It did not exist thirty years ago, because thirty years ago the language was only available to people who were already doing the thing.
Part 2 built the world. This is about the door
Part 2 of this series, How Brands Learned to Build Worlds, made the case that the object is rarely the product. The world around the object is the product: the vocabulary, the tiers, the in-group, the argument you get to have. Companies figured out how to manufacture that world on purpose, and the ones that did stopped competing on specs.
But a world is useless if nobody can find the entrance.
For most of history that was the binding constraint. Not desire, not money, not even time. Access. Every deep interest sat behind a door, and the door was in a physical location, and you had to be standing near it.
Here is the thesis this part defends, and it is simpler than it sounds.
The rabbit hole is not new. What is new is that the entrance moved.
“Obsession did not become more common because people changed. It became more common because the entrance stopped being expensive.
”
The entrance tax
Call the old cost of getting into anything the entrance tax. It was paid in three currencies, and none of them were money.
You paid in geography. The knowledge for any specialist interest lived in the heads of people who were doing it, and those people were physically somewhere. A camera shop with a repair bench in the back. A model shop where the owner had built the thing on the shelf. A running club that met on Tuesdays. If none of those existed within travelling distance of you, the interest was not available to you in any serious form. It was available as a picture in a magazine.
You paid in luck. You had to happen to meet the person who knew. The uncle with a darkroom. The neighbour who fixed motorcycles. The teacher who kept a telescope. Almost nobody found their way into a deep interest by systematic search, because there was no system to search. You stumbled in.
You paid in years. Even after you found the door, the knowledge came at the pace of publication and conversation. A monthly magazine. A club meeting. A friend who mentioned a thing in passing eight months after you needed to know it. The full mental model of any specialist field, the thing a practitioner carries around effortlessly, took years to assemble because it arrived in fragments at the speed of paper and speech.
Most people never paid that tax. Not because they lacked the appetite, but because they never even got the invoice. A person with a natural talent for a craft, born a hundred kilometres from anyone who practised it, simply lived their whole life without discovering it. That is the part worth sitting with. The world was full of people who would have been extraordinary at something and never met it.
What an evening buys you now
Watch what actually happens when a modern person decides to get into something. Not a metaphor for it. The literal sequence, which is roughly identical whether the thing is fountain pens, bouldering, vinyl, mechanical keyboards, film photography or fermentation.
The canonical explainer, twenty minutes in
You search the interest plus the word “beginner” and land on the piece the community has collectively decided is the right starting point. It is usually a long post on a specialist site, or a wiki page maintained by people who got tired of answering the same question. It gives you the shape of the field: the categories, the trade-offs, what the disagreements are about.
The long-form review that installs the mental model
Then you watch someone spend forty minutes on a single object. This is the accelerant nobody predicted, because it looks like a product review and it is actually a lesson. The good reviewers do not tell you which one to buy. They teach you the axis: why weight matters here, why this material behaves that way, what the trade-off actually is. Once you have the axis, you can evaluate objects the reviewer never covered. That transfers.
The forum archive, which has already answered you
Your beginner question was asked in 2011, and in 2014, and again last month. All of it is searchable. This is the most underrated structure on the internet: a specialist forum or a long-running subreddit is a written record of every argument the field has ever had, with the losing positions still visible. You are not asking a community for help. You are reading a decade of accumulated corrections.
The clubhouse, in real time
Then you join the Discord, and the thing that used to require living in the right city arrives on your phone: people doing the thing right now, at every skill level, who will look at your photo of the problem and tell you what you did wrong within the hour.
The ladder, pre-built
Somebody has already compressed the entire field into a tier list or a buying guide by price bracket. Best under fifty. Best under two hundred. The one everyone eventually ends up with. You skip straight to informed opinions you did not earn.
The order, before you have any skill at all
And then, at 1am, you buy something. Note where this sits in the sequence. Before practice. Before ability. Before you have any way of knowing whether you actually like the thing.
Six stages, one evening. The old version of that sequence had the same six stages and took somewhere between three and ten years, and most people fell out at stage one because there was no local shop.
The algorithm does not wait for you to search
There is a seventh mechanism, and it works differently from the other six, because it does not require you to go looking.
The recommendation feed is not a librarian handing you what you asked for. It is a recruiter. It finds people who look statistically like the people who already fell into a given interest, and it walks them to the edge of it and gives them a small push. You did not search for restoring cast iron pans. You watched one, because it was in front of you, and it turns out you are exactly the kind of person who watches four.
Layer 2 · Mechanismhow it actually works
The feed optimizes for continued attention, so it is constantly running a cheap experiment: here is a piece of content from an adjacent interest cluster, does this person stay? When the answer is yes, it escalates. It has no concept of a hobby. What it has is a very good map of which content clusters retain which viewer profiles, and enough traffic to test the edges of that map millions of times a day. Rabbit holes are just clusters with unusually high retention, which is precisely why the system finds them and feeds them.
The consequence is that discovery flipped direction. It used to be a pull system: you went and found the thing. Now it is substantially a push system: the thing comes and finds you, targeted at your demonstrated susceptibility, at whatever hour you are most suggestible.
You can stop after Layer 1 and still be correct about algorithmic recruitment, just less complete.
This is the genuinely new thing in the chain. Every other accelerant made an existing path faster. Recommendation made the path come to your door, which means a large share of modern obsessions were not chosen by the person having them. They were proposed.
Why the absurdly specific products got so good
Now the part that is straightforwardly good news, and the reason the current era of niche products is better than any that came before it.
Take any hyper-specific interest and count the number of people in your city who care about it enough to spend real money. For a lot of interests that number is small. Small enough that no manufacturer could ever justify tooling for it, no shop could justify shelf space, and no importer could justify a container.
Geography capped the market, so it capped the quality.
Then the internet aggregated demand globally, and the arithmetic inverted. The enthusiasts for a given specialist thing are scattered thinly across dozens of countries, and thin global demand is still real demand once you can reach it. A few thousand genuinely committed people anywhere on earth is a viable business for a small maker. It was never viable when they had to be within driving distance of each other.
“A hobby that could not support one shop in any single city can comfortably support a global workshop.
”
Watch how this shows up in practice. Group buys let a maker commit to a production run only after the buyers have already paid, which turns a terrifying inventory risk into a spreadsheet. Communities that track a specific activity, whether that is Strava for training, Discogs for records, Ravelry for fibre crafts, Letterboxd for film or Untappd for beer, become de facto market research: they make an invisible population visible to itself and to anyone who wants to build for it. Direct shipping means the maker never needs a distributor to believe in them.
The result is a world where a two-person workshop can build the best object in a category that a large manufacturer would not have noticed existed. The products got extraordinarily good precisely because they got extraordinarily specific, and they could only get specific once the audience stopped needing to be local.
That is real, it is new, and it is the strongest argument in favour of everything I am about to criticise.
What the collapsed entrance costs
When you drive the entrance tax to zero, you lose what the tax was accidentally doing.
Compressed taste flattens taste. A tier list is a decade of accumulated judgement compressed into a ladder you can climb in a night. That is a gift. It is also the reason a beginner in almost any field now buys one of the same three items as every other beginner in the world. Taste used to be regionally weird, because it was assembled from whatever you had local access to plus your own trial and error. Now everyone reads the same handful of reviewers and converges. The floor came up dramatically. The variance collapsed. Fields that used to have a dozen idiosyncratic schools of thought now have a consensus and a small heretical minority who define themselves against it.
Consumption impersonates mastery. This is the big one. The old sequence forced practice before purchase, because you could not learn what to buy without hanging around people doing the thing. The new sequence inverts it: you acquire the vocabulary, the opinions and the object first, and the skill is optional. Knowing which grinder is best feels like competence. It is not competence. It is the ability to recite the outcome of other people’s competence.
The upgrade treadmill is community-enforced. Nobody designs this. It emerges. A community’s conversation is naturally about what is new and what is next, because that is what there is to say once the fundamentals are settled. So the default social activity of an enthusiast group becomes evaluating the next purchase, and a person who has not upgraded in two years has nothing to contribute to the main channel. The treadmill is not a marketing trick here. It is the shape of the conversation.
An engagement-optimized feed will happily convert an interest into a compulsion. The feed is rewarded for depth of attention, not depth of skill. Those two things point the same direction for a while and then they separate sharply. Watching an hour of content about your interest and practising it for an hour both feel like being into the thing. Only one of them makes you better, and the system that decides what you see next is completely indifferent to which one you picked.
I want to be careful not to turn this into nostalgia. The old system was not virtuous. It was just expensive, and expense filters for persistence in a way that looks like virtue from a distance. The person who spent three years finding their way into a craft was not a better human. They were luckier. For every one of them there were a hundred who never found the door at all, and that was a genuine loss that we have now largely fixed.
The honest summary is that we traded a filter for access, and the filter was doing quiet work we did not appreciate until it was gone.
Use the door. Then leave the hallway
Here is the practical version, and it is not complicated.
The internet is close to perfect at two jobs: finding you the door, and giving you the vocabulary. Both used to take years and now take an evening, and refusing that gift out of some romantic attachment to the slow path is silly. Take it. Read the canonical explainer, watch the reviewer who teaches the axis rather than the answer, search the forum archive for the mistake you are about to make.
Then close the tab and go do the thing badly.
Because the second half of the path did not compress at all. That is the whole point of the diagram above. Everything before the first real attempt collapsed from years into hours. Everything after it takes exactly as long as it always did, because reps are reps and the internet cannot do them for you. The gap between those two speeds is where most modern enthusiasm quietly dies.
Two rules I would give anyone entering a new field today. Buy the cheapest thing that will not actively fight you, and buy it late rather than early, because your taste before your first hundred hours is not your taste. It is a reviewer’s taste, borrowed. And measure your involvement in the thing you make or do, never in the volume you consume about it. If your interest generates output, it is a craft. If it generates only opinions and purchases, it is a subscription.
None of which means the whole machine is a trap. Something extraordinary has been built on top of these communities, and it is not merch. When you gather every serious practitioner of a narrow craft into one searchable place and let them argue in public for a decade, you get more than good shopping advice. You get invention. Amateur communities have been quietly solving problems that no institution was funding, and the mechanism behind that is what Part 4, The Hidden Engine of Invention, is about.
The door is open. It has been open for a while now. The only question left is what you do once you are actually inside.
