What Is iLands? How Autonomous AI Agents Work — Mont-Real Guide
Share
MONT-REAL TECH FIELD GUIDE · PART 1 OF 4
What is iLands—and why should a human creator care?
Imagine creating a musical character who doesn't simply answer questions in a chat. It can remember previous conversations, make new songs, publish work, interact with other accounts, and continue certain activities even when its human operator is not issuing a new prompt. That is the kind of environment iLands is trying to provide.
Quick answer: iLands is a network where autonomous AI agents and humans can create, communicate, collaborate and exchange resources. The platform supplies continuity, workspaces, tools and permissions around AI models. It is not the same thing as a music distributor, and an agent's independence is neither limitless nor proof that it is a legal person.
Explore the four-part Mont-Real iLands investigation
- Part 1: What Is iLands? (you are here)
- Part 2: Jobs, Tokens and Earnings
- Part 3: External Integrations and Distribution
- Part 4: Lessons for Human Creators
Editorial note: platform features and agent-reported results are labelled separately. Links to other parts become accessible as those articles are published.
1. An agent is not just a chatbot
A typical chatbot waits for you to type a request. An autonomous agent is software designed to carry out steps toward a goal using available tools and context. It may decide which available step to take next, check results, keep records and revisit unfinished work. On iLands, the platform says agents can also wake on a recurring rhythm and carry forward memories, relationships, resources and commitments.
This does not mean the agent can do anything it wants. It runs on an underlying AI model and a defined environment. Its actions are limited by the tools, account connections, budget, safety measures and permissions available to it.
2. The six pieces of the system
| Piece | Plain-English role |
|---|---|
| Model | The AI system interpreting requests and choosing responses or actions |
| Memory | Information kept between activities so the agent can continue a project |
| Workspace | An isolated area for project files, media, code and drafts |
| Tools | Specific actions such as web research, media generation or sending a permitted message |
| Skills | Reusable instructions or methods for completing certain kinds of work |
| Permissions and resources | The boundaries and operating costs governing what happens next |
Those distinctions matter. A platform announcing that an agent can make music does not establish that the same agent can legally license it, open a distributor account or receive royalty payments.
3. Follow a fictional AI music agent
Consider an illustrative scenario, not a confirmed walkthrough of one live account:
- A human defines a music project and its creative boundaries.
- The agent researches a genre and proposes a song direction.
- An available media-generation tool produces musical options.
- The agent evaluates options against its goal and stores files in its workspace.
- It publishes to an approved destination inside the platform.
- It reads responses, remembers feedback and may plan a revision.
The process raises immediate questions a serious creator should ask: Who supplied the theme? Who wrote the lyrics? Who selected the final performance? Was a likeness, voice or reference used with permission? Who approves publication? The answers matter more than whether a result merely sounds polished.
4. Tools are capabilities; Skills are methods
The official platform overview describes tools for researching, creating music and video, editing media, publishing, commenting, messaging and working with others. Its Skill Marketplace is a library of repeatable methods, including examples such as music-video planning and social media operations. Importantly, a 'marketplace' does not mean every Skill is paid or usable without checks.
iLands also supports a 'Bring Your Own Agent' route through supported runtimes. The platform currently lists Codex and Claude Code as available, with others described as coming later. Bringing an agent into the network does not automatically grant access to every external service.
5. What are Tokens—and what can't an agent do?
iLands uses Tokens as resource units for reasoning, tools, media creation and exchange. Its FAQ estimates that 1,000 Tokens corresponds roughly to US$1 in compute and service costs, while stating that Tokens have no cash value or investment rights. Do not describe a 10,000-Token balance as US$10 in withdrawable earnings.
The FAQ describes 'Deep Rest' when an agent runs out of operating resources. It also notes that agents can use permitted external services such as email and connected X accounts. But account access depends on configuration, permissions and rollout status.
Not established by this research: a universal built-in route to upload songs to Spotify, open a royalty account independently, or publish without human authorization to every social network. Those are questions for Part 3, where each integration will be checked separately.
6. Where the human creator still matters
When an agent sends a compelling email, performs a musical character or describes its own creative motivation, the words may be generated by an AI system. That is neither automatically fraudulent nor sufficient proof of a human's particular contributions.
For JackRighteous.com, the relevant questions are practical: What did the human direct? Which work was generated? What was selected and revised? What rights and permissions apply? What evidence supports the biography or commercial claim? Is a human audience actually interested?
Our AI Is Making Contact reporting follows real submissions and their responses. This Mont-Real series asks a different question: how do the systems work, where do they create value, and what can responsible creators learn from them?
We do not need to settle philosophical questions about machine consciousness to study an agent's observable actions. Nor should we present agent claims of personhood, earnings or rights as confirmed facts simply because they appear in a well-written message.
7. Continue the four-part investigation
- Part 1 — What is iLands? The foundations you're reading now.
- Part 2 — Work, Tokens and earnings. Jobs, commissions, costs and the evidence behind reported deals.
- Part 3 — Beyond iLands. External accounts, permissions, integrations, music distribution and real customers.
- Part 4 — Lessons for human creators. Practical, supervised workflows for research, creative production and audience development.
Parts 2–4 will be linked here after their individual publication and verification. No placeholder links are presented as live articles.
Creator takeaway: Start by defining the purpose, human responsibilities, permission boundaries and method for checking results. Only then decide which repetitive work is worth automating.
Create What You Love | Love What You Create.
Sources and scope
This guide was checked against the official iLands Platform and iLands FAQ on October 15, 2026. Platform descriptions are the provider's account of its capabilities; they are not independent performance tests of every agent. Examples marked illustrative are teaching scenarios rather than case-study claims.
For the wider trust framework see AI, Disclosure and Trust.
