300万円の仕事が、5万円以下になったら何が起きるか

The answer is simple. The power dynamics between the “requester” and the “requested” will flip.

This week, two news stories caught my attention. One is about a successful project that utilized AI to natively port “Mario Kart Wii” to PC. The other is about the music production software “Reaper DAW” now being controllable by AI agents.

At first glance, these seem to be stories from two different genres: gaming and music. However, the essence is the same.

The cost of “jobs that could only be done by experts” has changed dramatically.

What does this mean for small and medium-sized enterprises? I want to delve deeper into that.

—

マリオカートWiiのPC移植——エミュレータではなく「ネイティブ移植」という衝撃

First, let’s clarify the facts.

In this project, the code for Mario Kart Wii compiled for the Wii was successfully recompiled for PC with the assistance of AI. This is entirely different from running it on an emulator. An emulator mimics the Wii’s hardware through software to “force it to run,” whereas this project has converted it into code that runs natively on a PC.

What were the results?

• Support for 4K resolution
• Removal of frame rate limitations (originally fixed at 30fps → now capable of over 60fps)
• Support for over 200 custom tracks
• Easy introduction of MODs

Traditionally, to achieve such a native port, a team of reverse engineering experts would need to work for several months to years. Outsourcing such a project would cost several million to tens of millions of yen. With the support of AI for code analysis and conversion, it has now become feasible for small teams.

The important point here is not merely that “Mario Kart can now be played.”

The cost of “transferring existing software assets to another platform” has dramatically decreased.

What does this mean when applied to small and medium-sized enterprises?

For example, consider an in-house business system created during the Windows XP era. It still works, but it’s outdated. The company that created it no longer exists. They want to port it, but when they got a quote, it was 5 million yen—this is a common story in local SMEs.

If the accuracy of AI-driven code analysis and conversion improves, the cost of porting such “legacy systems” could drop to one-tenth or even one-hundredth. A project that costs 5 million yen could become 500,000 yen or even 50,000 yen.

—

Reaper DAWのAI制御——「操作スキル」という参入障壁が消える

Next, let’s talk about music production.

Reaper DAW is a high-functionality music production software used by professional music producers. The personal license is inexpensive at $60 (about 9,000 yen), but mastering it requires a significant learning cost. Mixing, mastering, and adjusting effects—these are areas of “craftsmanship,” and outsourcing to a professional engineer can cost tens of thousands to hundreds of thousands of yen per song.

The recent announcement is that AI agents can now directly control Reaper. Users can give instructions in natural language like “bring the vocals a bit more forward” or “make the overall sound warmer,” and the AI will automatically adjust the parameters in Reaper.

What does this mean?

The value of “knowing how to operate the software” is rapidly approaching zero.

This is not limited to music production. The same phenomenon will occur with Photoshop, Excel, CAD software, accounting software—every specialized software will experience this. The structure of hiring or outsourcing to find “people who can use this software” will change.

Let’s consider this in the context of small and medium-sized enterprises.

“We’re paying 150,000 yen for outsourcing because we don’t have anyone who can edit videos.”
“We’re spending 30,000 yen on a flyer because we don’t have any employees who can use design tools.”

These costs could drop to less than one-tenth with AI agents controlling the software. Moreover, they would be available 24/7 and would not be dependent on specific individuals. Even if someone leaves, the system remains intact.

—

本当の論点——「専門スキルのコスト崩壊」の先に何があるか

The porting of Mario Kart and the AI control of Reaper may superficially seem like “interesting technology news.” However, if we look at the structure, they are saying the same thing.

“Jobs that could only be done by experts” will become tasks that can be performed by anyone with AI assistance.

Let’s organize this change with numbers.

Of course, this doesn’t mean that all of this will be realized today. But the direction is clear. Costs will drop dramatically.

And importantly, when costs drop dramatically, the criteria for deciding whether to do something changes.

A project that costs 3 million yen might lead to a discussion of “let’s consider it for next year’s budget,” but if it costs 50,000 yen, it becomes “let’s do it next week.” No need for lengthy approvals; the president can decide and implement it immediately.

This structure is overwhelmingly advantageous for small and medium-sized enterprises compared to large corporations. Large companies may take three months to implement a 50,000 yen tool due to security reviews, legal checks, and approval processes—the cost of decision-making is higher than the price of the tool itself.

In a small or medium-sized enterprise, if the president says, “Let’s use this,” they can start moving from tomorrow. This speed difference is the greatest weapon that SMEs have in the age of AI.

—

じゃあ、何から始めるか

Let’s leave the abstract discussions behind. What should we do specifically?

1. First, list the “outsourced jobs”

Flyer production, video editing, system modifications, data entry, translation—write down all the money you spend externally each month. Look for items among them that could be internalized with AI.

2. Experiment with outsourcing costs below 50,000 yen per month

There’s no need to dive straight into porting core systems. Start with small jobs. A single flyer, summarizing meeting minutes, a simple data aggregation tool. Begin with amounts that won’t hurt if you fail.

3. Create a “system” rather than relying on “people who can use it”

It doesn’t matter if only one employee can use the AI tool. Create a system where “if you follow this procedure, anyone can achieve the same results.” Standardize prompts, organize procedure manuals, and create output checklists. Avoid dependency on specific individuals.

4. Redirect the saved costs to “things only humans can do”

If AI saves you 200,000 yen a month, allocate that 200,000 yen to sales activities or customer interactions. Assign tasks that AI excels at to it, allowing humans to focus on “judgment” and “relationship building.”

—

まとめ——変化は「気づいたら終わっていた」の形で来る

The porting of Mario Kart Wii to PC and the AI control of Reaper may end up being news that many people think, “Oh, that’s interesting.”

However, what these two stories indicate is the fact that “the value of specialized skills is structurally decreasing.” This change will manifest one day as, “I suddenly realized that competitors had internalized with AI, and the cost structure was completely different.”

For local SMEs, this is an opportunity. Unlike large corporations, there’s no need for significant investments. What’s needed is simply the speed of judgment to “let’s try it first.”

In a world where 3 million yen becomes 50,000 yen, the first company to act will win. There’s no reason to wait any longer.