Version 1.1: images, agents, and learning on the web

The new image missions let you write a prompt and generate a real image without leaving mainiø. You can see how subject, setting, composition, and style change the output, then revise the prompt and try again.

That immediate loop matters. Image generation makes more sense when you can compare intention with output: what the model understood, what it invented, and which part of the prompt actually changed the result.

Image creation is free to try. mainiø+ opens the full available path and includes 30 generations each month for deeper practice.

Language models can work with more than text. In the new vision mission, Claude examines selected lesson images and responds to questions about what it sees.

The useful lesson is not that AI can name objects in a picture. It is learning where visual analysis helps, how confidently a model can describe something uncertain, and why a plausible answer still needs human judgment.

For now, the mission uses images provided inside the lesson rather than photos uploaded by learners. That keeps the first experience focused and gives everyone the same material to question and compare.

Agents are becoming a practical part of AI work, but the word is used loosely. The first six lessons build the foundation: what makes an agent different from a single prompt, how tools and feedback loops work, when delegation helps, and where autonomy creates risk.

Version 1.1.1 adds six more lessons built around work people can recognize. They cover research agents, recurring briefings, no-code automation, agent teams, and building a first website with agent assistance.

The path is designed for non-coders too. You do not need to build an agent framework to understand how to scope a task, choose the right level of autonomy, and stay responsible for the result.

The free version remains a real way to use mainiø. After the expanded first-day introduction, it includes three new lessons a day, selected live-AI experiences, and the core learning and progress system.

mainiø+ is for people who want more depth or a faster pace. It removes the daily lesson limit, opens every mainiø+ path and module, includes the personalized AI Skill Map, provides 30 image generations each month, and gives early access to new content.

That distinction is intentional. The free plan should be useful on its own. mainiø+ should make the experience broader and more personal, rather than fixing an artificially incomplete product.

The iPhone app remains the most complete mainiø experience, but it should not be the only place someone can start learning. This is version one of mainiø Academy: our first browser-based course, with 24 practical lessons covering AI basics, prompting, everyday work, critical judgment, repeatable workflows, and choosing the right tool.

Academy is free and does not require an account. Progress stays in your browser, so you can begin a lesson immediately and return on the same device later.

There are 24 lessons on the web for now, compared with more than 150 in the app. The web course is deliberately focused: it does not copy every app mechanic onto a larger screen. It offers clear reading, small exercises, and useful takeaways for people without an iPhone and for anyone who prefers a desktop learning environment.

The visible features came with less visible work: a refreshed Skill Map with an Agents axis, a full lesson accuracy pass, a smoother first lesson, clearer mission states, and fixes for sign-in and subscription reliability.

On compatible iPhones, one lesson can also use Apple Intelligence entirely on-device. The words in that interaction do not leave the phone. It is a small part of the update, but an important example of how different kinds of AI can fit into learning.

V1.1 is not a finished destination. It is the first version where mainiø can teach with live AI, not only teach about it, while the Academy begins extending the same practical approach beyond iPhone.

For now

The next question is not how many AI features can fit into the app. It is which experiences genuinely help someone form better judgment and use AI with more confidence.

For now, V1.1 gives us more ways to answer that question: create something, examine something, delegate something, and learn from what comes back.