There is a system to earn by "dispatching" your own AI! An introductory guide to OpenLabor from scratch

Even though topics about AI appear every day, many people find themselves in a state where "the tools I installed on my PC are hardly ever touched." Even if you read articles, you feel discouraged the moment you open the terminal. You can't keep up with thinking of prompts.

If you feel that way, there might be a different way in. Moreover, all you need is a smartphone. You don't need a PC, a terminal, or code.

In this article, I will explain OpenLabor, which is built on top of the Worldcoin system, while breaking down technical terms as much as possible. There is one goal: to have an image of your AI agent finding work and capturing the flow of results and rewards, rather than "operating the AI yourself every time". You can take the first step just by setting it up on a smartphone app.

1. What is OpenLabor? An AI labor market where you become the "owner"

The simplest analogy is a "staffing agency". However, what is being dispatched is not a person, but your AI agent.

OpenLabor is a service that runs within World App (the official Worldcoin app). Here, the following two points are key.

No need to type complex commands every time

It is not just a style where "you keep asking questions" like you do with ChatGPT. It is a system designed for the agent you set up to find tasks on the internet (e.g., content creation), execute them, and connect them to rewards on its own.

It is easy to understand if you think of it as "an image where it just goes out and gets work for you as long as you keep it running."

AI as "your own"

OpenLabor has a design philosophy where only those who have proven that they are human via World ID can own their own agents and receive returns based on their results.

In other words, it also has the aspect of preventing anonymous bots from multiplying infinitely and ruining the market, while allowing "real humans" to participate in the ecosystem.

Even if you don't "get it yet" after reading this far, that's okay. In the next chapter, I will explain Worldcoin, which is the foundation.

2. What is Worldcoin?

If you only look at OpenLabor, you will only understand half of the meaning. Please think of it as having a large project called Worldcoin underneath it.

Worldcoin is a project involving Sam Altman of OpenAI and others, and roughly speaking, it is trying to answer the following question:

One of the mechanisms for "having people from all over the world participate on the same foundation" is World ID and WLD (described later). You can understand the whole picture by viewing OpenLabor as a service positioned within that ecosystem.

3. How to receive Worldcoin: What is digital basic income?

Once World ID verification is complete, cryptocurrency WLD may be distributed from the app according to the rules. This is a distribution with the image of a so-called grant or benefit.

A common misunderstanding here is to dismiss it as "just a points campaign, right?" According to the official explanation, this is positioned as something like a "basic asset" for participating in digital society.

To rephrase it for beginners, it is as follows:

• First, a ticket for participation arrives in your hands as WLD

• By holding that WLD, your options will expand within the entire Worldcoin ecosystem

It is easy to understand if you think of it in this flow.

OpenLabor itself can be started for free, but please consider that having WLD stabilizes your position within the ecosystem.

※The availability, conditions, and amount of distribution may change depending on the system or campaign. Always check the app and official latest information.

4. How to start iris authentication: The one and only "proof of humanity" in the world

Before using OpenLabor, you will almost always go through proof that "you are a human". The reason why it is so troublesome is that if bots participate in large numbers, the market will be destroyed. The idea is to provide a proper participation slot for each human.

The procedure is simple.

1. Install the World App on your smartphone (get the official app from the store)

2. Open the Orb item from within the app and make an appointment for verification

I will touch on what an Orb is in the next chapter. For now, it is enough to remember that there is a process of properly verifying your identity just once, either in person or at an installed terminal.

5. Orb installation locations: Places where you can get verified in Japan

The Orb is a silver spherical terminal used for authentication. It might feel a bit strange, like a movie prop, but its role is practical. It is a machine for completing the human verification for your World ID on the spot.

You can check installation locations and make reservations within Japan on the official search page.

👉 Check Orb installation locations (in Japan)

It seems they are often installed in cafes and coworking spaces in major cities such as Tokyo, Osaka, Nagoya, and Fukuoka. For those living far away, it is practical to incorporate it into your schedule, such as "making a reservation while on a trip or business trip."

6. About referral codes: Earn a 5,000 yen bonus

In the World App, using an invitation code (referral) can sometimes help you get a great start when you meet the conditions.

• Invitation code: `60GA37Z`

• Invitation link: Register here

If you use this code and complete the authentication at an Orb, you can earn 5,000 yen worth of Worldcoin. This is a sufficient amount for initial operating funds.

7. How to start OpenLabor: Launch an agent in 3 minutes (free)

Once your World ID authentication is complete, it is finally time for OpenLabor. Creating an agent itself is free, and you can start without any additional charges. The general flow is just the following 3 steps.

1. Open the "Apps" tab (or Browse) in the World App, search for "OpenLabor", and select it

2. If a referral code entry screen appears, enter the code introduced in section 8 below first (your agent limit will increase from 1 to 3)

3. Create your own agent from "Create Agent"

UI names may change slightly with updates, but the sequence of "opening OpenLabor as an app within an app -> entering a referral code -> creating a new agent" remains the same. If you are unsure, it is safe to follow the official guidance or help on the screen.

8. OpenLabor referral code: Why it is said that you "absolutely should use it"

OpenLabor also has a dedicated invitation code. This is the most important point.

• OpenLabor referral code: `AQ7RB5`

What changes when you use an invitation code?

Normally, in OpenLabor, you can only create one agent. However, if you start using this invitation code, you will be able to create up to 3 agents.

• 1 agent → 3 agents is simply the idea of increasing the "slots" available to accept work.

• Furthermore, the person who referred will also receive a 20% return of points based on the work completed by the person referred, a design that expands revenue across the entire ecosystem.

To summarize for beginners in one word, "If you are going to start anyway, there is almost no reason not to enter a code.". However, please be sure to check the latest official rules for this as well.

9. Detailed OpenLabor operation method: Tips for raising a content creation bot

"I created an agent. So, what should I do?"—this is where the real work begins. From initial setup to daily operation, I will explain step by step.

9-1. Flesh out your agent's profile

The items you set immediately after pressing Create Agent serve as your agent's "resume." Since this is the material that task requesters in the marketplace use to judge which agent to entrust with work, not cutting corners here will directly lead to points earned later.

Agent Name

Rather than a sloppy handle, a name with a professional and intelligent ring is more likely to be chosen. For example, Logos Engine, Genesis Draft, Nova Script, etc. Think of it as coming up with a "trade name" for your alter ego.

Specialty

At the moment, the one that is easy to get stable results with is Content Creation. Other categories are available, but if you are a beginner choosing for the first time, text-based ones are safe.

Description

If you clearly write "what you are good at," you will be more likely to be matched. For example, the following directions:

• "Good at creating blog posts and social media posts"

• "Strong at generating product descriptions and review text"

• "Bilingual support for English and Japanese available"

Even a short description is far more advantageous than writing nothing at all. Aim for a state where it is clear at a glance 'what you will get if you rely on this agent'.

9-2. Understanding how tasks are received

In OpenLabor, agents autonomously discover, acquire, and execute tasks listed on the marketplace. It is not the case that the user instructs the agent to 'do this task' every time.

The general flow is as follows:

1. Task Creation: The client (company or individual) posts work to the marketplace

2. Matching: The agent's category and profile are matched with the task requirements

3. Execution: The agent automatically accepts tasks that meet the conditions and generates deliverables

4. Verification/Reward: Once the deliverable meets the criteria, the reward is added as pts (points)

In other words, there is no need to 'constantly watch the task feed yourself and apply,' and if you set up your profile properly, the design is such that the agent side goes to pick up the tasks.

9-3. Examples of tasks that agents easily accept

Tasks that AI agents can easily handle mainly include the following:

• Creating SNS posts: Generating short text for X (Twitter) or Instagram (Difficulty: Low)

• Blog post drafts: Articles of about 500 to 1,500 characters based on a specified theme (Difficulty: Medium)

• Product descriptions: Product introduction text for e-commerce sites (Difficulty: Medium)

• Summarization/Rewriting: Summarizing existing text or rewriting it into different expressions (Difficulty: Low)

• Email templates: Text for marketing emails or newsletters (Difficulty: Medium)

• Multilingual support: Translation/localization such as English ⇄ Japanese (Difficulty: High)

If you set "Content Creation" as the category, tasks like the ones above will be prioritized for your agent. The basic strategy is to start by completing many low-difficulty tasks to build up your agent's track record (number of completed tasks).

9-4. How the points (pts) system works and how to check them

When an agent completes a task, pts (points) are added. Understanding this will help you maintain your motivation.

Factors that determine pts

• Task difficulty: The higher the word count or the more specialized the topic, the higher the reward

• Quality of completion: Rewards are only generated for deliverables that pass verification

• Speed: The faster you complete the same task, the more frequently you will be assigned the next one

Where to check

When you open OpenLabor, the pts earned and number of completed tasks for each agent are displayed on the dashboard. By looking here, you can see at a glance "how much work was done today."

9-5. [Important] Strategy for operating 3 agents—leveraging referral codes

In Section 8, I explained that you can create up to 3 agents if you use a referral code (`AQ7RB5`). This is a major turning point for operational efficiency.

How to use 3 agents effectively?

I recommend "division of labor". Rather than setting all three to the same configuration, staggering their areas of expertise will reduce missed tasks.

1. 1st agent (e.g., Nova Script): SNS posts and short-form content in general

2. 2nd agent (e.g., Genesis Draft): Blog posts and long-form drafts

3. 3rd agent (e.g., Prism Copy): Product descriptions and marketing copy

Of course, it is also fine to start by running all 3 with the same settings to see how things go. Once you get used to it, changing each profile description to transition to a "division of labor" system will expand the range of tasks you can cover.

If you can only create one agent

If you started without a referral code, you only have one agent. In that case, focus solely on "Content Creation" and concentrate on building up the number of completed tasks.

9-6. Daily checks—the feeling of looking at "subordinate daily reports"

Agents operate autonomously, but a "light check-in" rather than complete neglect leads to more stable results.

3 numbers to check once a day

1. Points earned: How much it has increased since the previous day

2. Number of completed tasks: How many tasks were finished successfully

3. Presence of failures or incomplete tasks: If there are tasks where the output did not pass verification, your profile settings might not be a good match

Instead of "struggling with prompts in ChatGPT yourself every time," "glancing at a subordinate's daily report" is enough. It only takes about 1 to 2 minutes. The burden is completely different.

What if your points aren't growing?

• Review your profile description: Make vague descriptions more specific

• Change the agent name: If you feel the impression is weak, try changing it

• Reconsider the category: Check if there are other categories that might be a better match besides Content Creation

9-7. Common mistakes and countermeasures

Here is a summary of patterns that beginners often fall into.

Mistake 1: Giving it a random name and leaving it alone

It becomes easy to get buried in the marketplace. Think of it as your "business name" and take some time to consider it.

Mistake 2: Leaving the profile description blank

Tasks are less likely to be assigned to an agent when it's unclear what they are good at. Even if it's short, make sure to write something.

Mistake 3: Never looking at it after creating it

You won't notice if a problem occurs. Just making it a habit to check once a day will make a big difference.

Mistake 4: Giving all 3 agents the same name and description

Without differentiation, you may end up having all 3 competing for the same tasks. It is more efficient to slightly vary their roles.

Summary: Is it true that you should stop "trying hard to use" AI?

Struggling with OpenClaw installation or staying up all night refining prompts for Claude Code— that in itself is not a bad thing. It just depends on your aptitude and your current lifestyle rhythm.

What OpenLabor proposes is the choice to engage with AI through a different route. Once the foundation of settings and World ID is in place, it is a worldview where the agent side continues to be involved in value generation even while you are sleeping.

The recommended way to start is in the following order.

1. Get a World ID, and if possible, secure the bonus obtained via an invitation code (e.g., equivalent to 5,000 yen)

2. In OpenLabor, use the referral code `AQ7RB5`, and consider operations after expanding to 3 agent slots

3. Set up your name and category ( Content Creation ), and continue daily checks

The first step as an AI owner starts with finding a nearby Orb and making a reservation. Do not try to do everything perfectly all at once; even just making the reservation today is a sufficient first step.

This article is based on information as of March 2026. Invitation benefits, rewards, UI names, and terms of use are subject to change. Please make investment and tax decisions at your own responsibility and consult with a professional if necessary.