You Can Do It Too: AI, Japan.co.jp, and the door the gatekeepers forgot

The door opens

Every generation gets one door. Most people walk past it because it looks strange, unfinished, or slightly embarrassing. In the early 1990s, that door was the World Wide Web. It did not arrive wearing a tuxedo. It arrived looking like academic plumbing. Blue links. Plain pages. Servers. HTML. A blinking cursor and the suspicion that something enormous had just been given away for free.

That is exactly what happened. Tim Berners-Lee invented the Web at CERN in 1989 to let scientists share information across institutions. In 1993, CERN put the Web software into the public domain and later released it under an open licence. That decision mattered because it meant no king, no tollbooth, no licensing priesthood. The Web could spread because nobody could stand at the gate and say: “Lovely idea. Please pay us rent forever.”

For a while, the field really did flatten. A kid, a garage business, a weird magazine, a tiny solar company, a Japanese-language hobby site, a local restaurant, a travel writer, a programmer, a poet, a troublemaker — all could publish. Not equally, not perfectly, but directly. You did not need permission from a newspaper chain, television network, retail shelf, distributor, booking agent or industry club. You needed a domain name, a server, patience and nerve.

Then the old guard noticed

At first, the old guard laughed. Then it ignored the Web. Then it bought websites. Then it panicked. Then it decided the best way to survive being leapfrogged was to rebuild the gates in a new shape.

The AOL-Time Warner merger, announced in January 2000, remains the great cartoon panel of that moment. The internet company and the old media empire were supposed to become the future together. Instead, the deal became a historic warning about timing, hubris and trying to turn a wild open network into a comfortable walled kingdom. When the bubble burst, the combined company took enormous losses, and the story became a business-school monument to “maybe the future should not be managed entirely by people protecting yesterday.”

But do not mistake that failure for the end of gatekeeping. The gatekeepers learned. The new walls were smarter. Search ranking. App stores. Social feeds. Platform rules. Algorithmic visibility. Ad marketplaces. Paywalls. Preferred partners. Bundles. Black-box moderation. Audience capture. Data capture. “Community guidelines” that sometimes looked suspiciously like “the house always wins.”

The open Web did not disappear, but attention moved onto platforms. The field that once felt like a public road became a series of malls. You could still set up a shop, but the landlord controlled the hallway, the signs, the rent, the security guard, the escalator, and the map that told customers whether you existed.

The cartel is not always a smoke-filled room

When people hear “cartel,” they imagine cigar smoke, secret handshakes and men in heavy suits laughing over steak. Sometimes power is less theatrical. Sometimes the cartel is just a set of incentives that all point the same way. Big platforms prefer scale. Big advertisers prefer certainty. Big incumbents prefer rules they can afford to navigate. Big compliance departments prefer systems that small competitors cannot survive.

The result can feel like a cartel even when nobody holds a villain meeting. The old players do not have to say “no new entrants” out loud. They can simply build a world where new entrants must buy traffic from the gate, obey opaque rules from the gate, give data to the gate, format themselves for the gate, and thank the gate for the opportunity.

This is why the Web’s original dream matters. Not because the 1990s were perfect. They were not. The early Web had scams, ugliness, bad design, broken links, and the kind of blinking text that should have been regulated by international treaty. But it had one great democratic instinct: publish first, ask permission never.

AI is the second door

Now another door is opening. This time it is called AI. It is not just a chatbot. It is a cheap research assistant, junior programmer, copy editor, design partner, translator, analyst, spreadsheet whisperer, customer-support draft writer, image brainstormer, code explainer, grant-writing assistant, marketing department, and occasionally a confident intern who must be supervised carefully because it just invented a source with the smile of a golden retriever.

That last warning is real. AI is not magic. It hallucinates. It can flatten taste into sameness. It can make lazy people lazier and noisy people noisier. It can tempt small businesses into publishing twelve bland paragraphs when one honest sentence would do. It can put bad advice in a clean suit.

But that is not the whole story. The whole story is that tools once reserved for large organizations are now reachable by tiny ones. Research that required a staffer. Design drafts that required an agency. Code prototypes that required a developer. Translation that required a vendor. Market scans that required a consultant. Customer emails that required a team. The small player can now start.

Not finish perfectly. Start. Starting is the revolution.

The small team is dangerous again

A big corporation has resources. It also has meetings about meetings. It has brand guidelines, procurement cycles, legal review, steering committees, stakeholder alignment, a deck about the deck, and a director of strategic acceleration who is currently unavailable until Thursday.

A small builder has less money, fewer people and more terrifying mornings. But with AI, the small builder can compress the distance between idea and prototype. That is the dangerous part. A one-person shop can test a landing page. A local publisher can build a bilingual edition. A solar contractor can draft a permit letter. A teacher can make learning materials. A startup can compare competitors. A photographer can plan a campaign. A writer can research a history, outline a piece, translate it, and publish before the committee has named the committee.

This does not mean effort goes away. It means the first wall gets lower. You still need judgment, taste, ethics, courage, persistence, and the ability to say, “No, that paragraph sounds like a committee trapped inside a toaster.” But the distance between wanting to build and beginning to build is shorter than it has been in years.

Japan should understand this

Japan has a special relationship with tools, craft and process. The country knows that small workshops can create great things. It knows that precision matters. It knows that a clever tool does not replace discipline; it extends discipline. That is why AI should not be seen only as a threat to white-collar workers or a toy for Silicon Valley.

For Japan’s small businesses, creators, local publishers, tourism operators, restaurants, manufacturers, farmers, engineers and one-person companies, AI can become a practical tool. Write the English page. Draft the export email. Summarize the regulation. Make the flyer. Translate the menu. Prototype the app. Build the FAQ. Clean the spreadsheet. Plan the video. Explain the tax form. Ask the dumb question privately before asking the expensive expert publicly.

That is not fantasy. That is Tuesday.

The old story said small players should stay in their lane. AI says: make a new lane, put up a sign, and see who drives down it.

Japan’s polite gate: one company, one .co.jp

Japan has its own version of the old gate, and it wears a very respectable suit. The .co.jp domain is one of the most trusted commercial signals on the Japanese internet. It says, “This is a real company in Japan.” That trust has value. But the same system also carries a peculiar restriction: organizational/geographic JP domain names, including .co.jp, are limited to one registration per registrant.

The history deserves context. JPNIC’s own timeline explains that the old one-domain-per-organization rule helped deter disputes, including domain registrations for resale. General-use .jp later removed that limitation, but the organizational-type structure kept its one-per-registrant discipline. In other words, what began as order and anti-squatting hygiene now also functions as scarcity policy for serious Japanese commercial identity.

That is Japan at its most frustrating: order so proud of itself that it forgets creativity also needs oxygen. A company can run multiple projects, serve multiple audiences, build multiple publications, and create multiple brands. The market does not ask whether your administrative drawer has enough corporate registration certificates. The market asks whether you can serve people.

This is why Japan.co.jp matters. It should not exist as a sleepy trophy domain. It should be used as a crowbar.

Japan.co.jp is the proof

Japan.co.jp is not a giant media corporation. It is not a tower full of editors, translators, designers, researchers, data teams, audience-development departments and executive retreats where people say “synergy” near pastries. It is a single entrepreneur using AI, history, stubbornness, affection for Japan, and a frankly unreasonable amount of caffeine to build a daily bilingual newspaper.

That should make the old guard nervous. Not because Japan.co.jp is perfect. It is not. It is alive. It learns in public. It ships. It improves. It publishes deeply, visually, bilingually and daily in a way that, only a few years ago, would have required a newsroom and a budget large enough to make a banker blink.

The message is not “look at us.” The message is “look at what is now possible.” If one entrepreneur can build Japan.co.jp into a daily visual newspaper, then a regional craftsman can explain his work to the world. A small inn can publish in English. A solar contractor can compete with national marketing departments. A local museum can tell its story. A teacher can build curriculum. A town can make itself visible. A weird little business can become less weird and more found.

The gate is not the world. The gate is just furniture.

But do not rent your soul back to the gatekeepers

There is a trap. The same old guard that built the second-generation gates around the Web would love to build the gates around AI. The pattern is already visible: closed models, platform lock-in, data moats, distribution control, default settings, enterprise bundles, and app ecosystems that make independence feel convenient until it is too late.

Use the tools, but keep your house. Own your domain. Keep your email list. Save your files. Build direct relationships. Publish on your own site. Use platforms for distribution, not identity. A platform can be a megaphone. It should not be your only address.

The Web’s first lesson was: open standards matter. AI’s first lesson should be: portable capability matters. Do not let the new door become another rented hallway.

The great equalizer is not the machine. It is the builder.

AI does not make every small business great. It will not save a bad product, dishonest promise, lazy service or empty idea. The machine can help you write. It cannot make you care. It can help you design. It cannot give you taste. It can draft a business plan. It cannot force you to call the customer, fix the problem, ship the product, or admit you were wrong.

That is good news. If AI did everything, everyone would be equal in the worst way: equally bland. The opportunity is not that the machine replaces human difference. The opportunity is that it removes some of the drudgery that used to smother human difference before it reached the market.

AI can help the honest builder show up. It can help the local expert publish. It can help the immigrant founder translate. It can help the small manufacturer explain. It can help the weird artist prototype. It can help the independent newspaper compete with the glossy giant that has a bigger office and fewer original thoughts.

You can do it too

That sentence is dangerous. It always has been. The old guard hates it because it breaks the spell. Every gatekeeper survives by convincing the crowd that the gate is natural. “This is just how publishing works.” “This is just how distribution works.” “This is just how marketing works.” “This is just how software works.” “This is just how influence works.”

No. It is how yesterday’s winners arranged the furniture.

The Web moved the furniture once. AI is moving it again. The room is messy. Some of the new tools are overhyped. Some people will use them badly. Some old players will capture parts of the stack. Some new players will become old players faster than they should. That is all true.

Still: the door is open.

Build the page. Launch the service. Write the story. Translate the idea. Make the prototype. Call the customer. Publish the awkward first version. Improve it tomorrow. Do not wait for the old guard to bless you. They are busy polishing the locks.

You can do it too. Not because AI will do it for you, but because AI can help you begin before the cartel notices you have already walked through the door.