The Only Winning Strategy in the AI Era

Many people are likely frustrated, thinking, "I thought I could make money easily using AI, but it's not working at all."

Or perhaps many are wondering, "What does it even mean to make money with AI in the first place?"

Those people are misjudging the essence of the AI era.

What is the biggest difference between the AI era and previous eras?

In a word, it is the speed at which markets saturate.

In the pre-internet era, information distribution was slow and lacked immediacy. In a world where network effects didn't work as much, information circulated within closed networks, and the information you could get differed depending on where and when you were born. Making money from that information asymmetry was the mainstream.

Since the internet revolution, that gap has steadily narrowed. But it still existed. With the spread of social media, it became even shorter. But there was still a grace period.

Then AI arrived.

AI is essentially an entity that provides the same answer to everyone. If you ask AI for a "good side hustle," it returns the same answer. If many influencers broadcast information that "this is profitable," everyone does it. People concentrate there intensely, the market saturates, and it is destroyed. This cycle runs at ultra-high speed.

This is the essence of the AI era.

It Doesn't Even Last Two Months

OpenClaw proved exactly that.

It became popular in February. By April, it was already obsolete. Search volume is less than 1/100th of its peak. A massive number of Mac minis are being sold on the used market. It didn't even last two months.

Markets that used to last about two years now end in one or two months. The rise and fall happen at an incredible speed.

If someone had written a book intending to make money with OpenClaw, they'd be finished. By the time it's published, the market is gone.

It's a terrifying sense of speed.

Don't Jump on Specifics

In other words, if you take a position based on specifics, you die quickly.

Specific know-how becomes obsolete immediately. Anything that fits a template, like "use this tool to make this," is usually like that. To put it simply, things like "How to make money with OpenClaw" or "You can mass-produce videos and slides with NotebookLM, let's make money with this" will quickly fade. You can only make money during the window before the information circulates, and that speed is increasing now.

This is where it becomes paradoxical.

Everyone seeks specific information. They want to jump on specifics like "this is profitable." But it's better not to jump on those specifics too much.

So, what should you not jump on? Things that everyone thinks, "That's certainly true."

Anything that makes you intuitively think, "This could definitely work," is no good. Because everyone else thinks the same. What everyone is saying is 100% no good. It was true in the past, but it's especially true now.

You feel safe because everyone is saying it. As human intuition, that's normal. But OpenClaw proved exactly that. Everyone said, "This is good, you can make money with this."

Now, just saying "OpenClaw" makes you feel incredibly behind. People say, "You can do that with Codex or Claude Code."

Engineers knew from the start. There was a cold perspective of "Is there even a scenario to use OpenClaw?" That perspective has now properly spread to the general public. That cycle is incredibly short.

And are the people who were pushing that really trustworthy? They'll just push something else next and call the previous thing "dead." Next, they'll say it's Substack or something. They're just trying to ride the trend; they are noise or pulses.

In short, the AI era can be called the era of paradox.

It Will Be an Era of Paradox

As the speed of market collapse increases, it is, in a sense, normalized. Because the speed of normalization is faster, you cannot make money properly unless you go to a more essential place. Even the ability to make money during a time lag is disappearing.

So, what should you jump on? Things that go against your intuition. Things that make you think, "Wait, really? I've never heard of that." However, it shouldn't be completely crazy either.

What everyone is saying → No good

Crazy stories (like 100x return just by putting in 1 million) → Out of the question

The subtle balance in between → This is the only place where a market remains

This balance is incredibly difficult. But that's the only place where a market is left. There's no room for expansion elsewhere.

When looking at people, look for consistency. Someone who said "you can make money with OpenClaw" can be cross-checked immediately. Usually, that person will push Claude Code next, and then something else after that. They're just betting on major things to ride the trend. Don't be misled by that. See if there is consistency in what they post.

This is the same for those who broadcast information. Even if it's not the case now, in a year or two, literacy will rise, and such broadcasters will be seen through.

The Fate of Homogenization

AI is a general-purpose technology. The more it evolves, the more versatile it becomes. Dynamics opposite to the SaaS era are at work.

SaaS was a model where various companies made various tools, and you used them in a fragmented way. AI won't be like that. Ultimately, it will converge into one or two, and others will be eliminated. It won't be anything else.

Most "vibe coding" tools have already almost disappeared. Replit and Lovable are also in a tough spot. Claude Code or Codex is fine. You can use something dozens of times better for about the same price.

Creative work is the same. It's a battle of who releases Seedance 2.0 first. Once agent functions come out, every company will follow suit. The same goes for the MCP trend. Many new tools and services will come out, but ultimately they will all become nearly the same. They will homogenize.

It will become "this is all I need." This is the fate of AI.

The Era Where AI Scores 80

A year ago, the common wisdom was "AI can only score 60, so humans must polish it." Now it's really scoring 80. Some things even score 120. Movies are truly insane. In a year, 90% of the screen might be AI-made.

AI is raising the average score. Everyone can produce quality of 80 or higher. When that happens, the gap in original superiority narrows too much, and from the recipient's perspective, there isn't much difference.

That's why you can't compete with cheap tricks. Even in broadcasting, there's not much point in sharing superficial tips.

The Era of the Salaryman is Ending

As generalization by AI progresses, it will replace labor. OpenAI announced the full-scale rollout of AI implementation support for companies. This is almost equivalent to a declaration that "we will eliminate employees." It's a given that consultants will die out, and employee reductions will happen normally.

News also came out that freelancers are returning to being employees. It's a simple dynamic: they can't make a living because their freelance work was replaced by AI, so they get a job. But even if they become employees, they'll just be laid off again, so it's meaningless.

The 100-year era of the salaryman is ending.

From now on, you have no choice but to operate with a merchant mindset. It's not the time to say you don't want to do this or that. You just have to do it. You have to do it, and AI will support everything.

Conversely, there are many good aspects. Because AI supports you, you can turn what you want to do—things closer to the essence—into a business. Even without initial capital, you can create services and products that can compete with major corporations. I want to go to that side. That is the ideal.

Opportunities Lie in What is Counter-Intuitive

There are actually many people who, at first glance, don't seem to be doing much but are making a lot of money. The things people do that make you wonder "Why does this person have so much money?" are things you've never heard of.

People making money with methods often talked about on social media have a low ceiling. Because it's something many people are doing.

"I've never heard of it, but the numbers are there." That kind of area is the best. The more counter-intuitive, the better. Because everyone else won't enter.

However, there are things that work in the long term and things that only work in the short term. You need to discern that, and the criteria for discerning are what I've talked about in this article.

Concrete Example: The 100 Billion Yen Blue Ocean of Minecraft Servers

After hearing all this, many people probably thought, "I understand the concept, but what specifically should I do?" So, I'll give one concrete example that I am currently working on.

The Minecraft server market.

The third-party Minecraft server market is worth 100 billion yen annually. Including the base game, it's hundreds of billions. It's a title acquired by the mighty Microsoft and is among the top 10 game sales worldwide.

And Japan is wide open.

Top overseas servers, like Hypixel, have annual sales of around 10 billion yen. Since it's digital content, the profit margin exceeds 90%. An annual profit of 9 billion. That's at the level of a top-tier listed company.

The mechanism is this: Minecraft servers are like MMORPGs where you can play multiplayer. You can enter for free, but they sell things like rank systems and skins—things that don't directly give a gameplay advantage—for a fee. Mojang (the developer of Minecraft) also permits this format.

"Can you make money with Minecraft?" "What even is Minecraft?"—The moment these reactions come back, it satisfies the "counter-intuitive" condition I mentioned earlier. You've never heard of it, and no one is talking about it. But you can confirm the existence of the market with data.

Why is Japan Open?

There are no businesspeople in the Japanese Minecraft community. This is the biggest reason.

There are users playing. There are a few people trying to make money as a hobby. But there's almost no one doing it with a proper business perspective. So, if you ask a Japanese Minecraft server operator, they'll say, "You can't make money with Minecraft servers." That's because they aren't doing it as a proper business. If you can do it with a business perspective, there's a huge chance of winning.

Clip videos of major overseas servers also get huge views in Japan. This means there is demand. People think it looks fun. Yet, there's no path to properly connect them. No one is doing it.

In fact, someone doing clips for DonutSMP (a major overseas server) has made about 20 million yen in a little over six months. 2 to 3 million yen a month. Even at a low estimate, 1 to 1.5 million yen a month. They haven't even put out that many videos. If it goes that far, it's incredibly good.

What AI Changed

And what's different from previous businesses is that you can operate a server with Claude Code or Codex.

We've entered an era where AI supports creating popular content like Skyblock (a genre where you gather resources while traveling between floating islands). Even without programming knowledge, the technical hurdles of server management have been drastically lowered by AI.

Treasures of the AI era lie in these "places where no one intuitively thinks they can make money" or "places that look normally difficult."

Probably not even 1 in 1,000 people who read this article will take action.

Difference from Roblox

Some people might imagine it's similar to making money by releasing games on Roblox.

There are similarities. However, on Roblox, being recommended on the platform leads to the most sales, so there's the hurdle of having to make games that Roblox likes. It's not your own server, so it's hard to control. And trends are fast. Competition is fierce. Because people know they can make money, it's a red ocean.

Minecraft has a market size not much different from Roblox, but in Japan, it is an overwhelming blue ocean. No competitors means it's easy to win. It's growing overseas, and there are no players doing it in Japan. It's the optimal solution in terms of time-machine management.

Creating Vertical Integration Yourself

Another important point is that there was no player to integrate elements that exist separately.

In Roblox, there are small businesses that provide tools to measure from upstream to downstream. But there wasn't for Minecraft. There was no product that consistently supported everything from advertising to conversion.

The world of the Web and the world of the Minecraft protocol are different. Therefore, you couldn't consistently get data on which video someone watched, how they were referred, and what items or ranks they bought within the server.

I made a plugin to connect those. Now, all marketing data from upstream to downstream can be captured.

Surprisingly, there really was no one doing that. Even though it's a proper business layer.

Companies that have end-to-end integration are strong. It's the same dynamic as Netflix becoming even stronger. How do you bite into the flow, or do you create that flow yourself? What I'm trying to do this time is like a small version of that.

Summary

Jumping on specific tools or methods is nothing but a risk in this era. Since the speed of market saturation is increasing, the era of making money through information asymmetry is over.

Things that everyone thinks "will work" die quickly. Find a market that is counter-intuitive but whose existence can be confirmed with data. Watch the posts of people with consistency. Take a position in a place close to the essence.

It's a difficult era. It's difficult for both those who broadcast and those who receive. But for those who act while accurately recognizing that difficulty, there has never been an era so full of opportunity.

If you're interested in content on how to conquer the Minecraft business, you should check it out here. I've posted many concrete examples, so even the free part should be educational.

You can view it from the link in the OpenChat below.

https://line.me/ti/g2/iouYfYhheBXkOKfz21bha8glm6kfkct0NL2OWA?utm_source=invitation&utm_medium=link_copy&utm_campaign=default

By the way, I'm talking about it in detail on Space, so if you want to know more, you should listen to that too.

https://x.com/satori_sz9/status/2054184261176221950?s=46&t=2qnM1Xf49MYeVh_XCTnslg