Anthropic Introduces Two Ways For Users To Learn In Its Claude Code And Claude App

The momentum behind AI has picked up so much speed that it’s becoming nearly unstoppable.

Once people realized that large language models (LLMs) possess remarkable logical and problem-solving abilities, programming naturally emerged as one of the first applications to explore. Anthropic, founded in 2021, was built with a mission to develop AI grounded in strong safety principles. Its unique “Constitutional AI” framework emphasizes honesty, reliability, and long-term societal benefit.

Claude, released during the period when OpenAI’s ChatGPT was dominating the AI conversation, was Anthropic’s direct response.

It was designed not just to match ChatGPT’s capabilities but to offer more consistent and safer interactions.

By May 2025, Claude 4 arrived with significant improvements in reasoning and coding capabilities. Building on this foundation, Anthropic now introduces two new ways for users to learn coding directly through Claude.

Amid the growing demand for programming skills and the rapid evolution of software development, a new practice called vibe coding is emerging. With this approach, a developer relies heavily on an AI to generate and refine code through natural language prompts.

In this new landscape, AI company Anthropic has a unique goal for Claude.

The company wants Claude to go beyond the role of a simple coding assistant and also serve as a tutor, helping users not only get the code they need but also truly understand what they are building.

Anthropic’s update introduces two distinct learning modes, designed to transform Claude from a simple chatbot into a true mentor.

The first is found in the Claude app.

Here, the new “Learning” mode, which was previously limited to Claude for Education users, is now available to everyone.

In this mode, Claude takes a Socratic approach, asking questions, guiding users through problems, and encouraging deeper thinking rather than simply providing answers. It’s an approach that helps users not only reach a solution but also understand the reasoning behind it.

The second is in Claude Code, where Anthropic’s developer-focused command-line interface.

In this environment, two preset /output-style modes have been introduced: Explanatory and Learning.

Explanatory mode walks users through the logic behind code decisions, discussing trade-offs and best practices as if a senior engineer were mentoring them in real time. Learning mode, on the other hand, pauses the coding process at certain points, leaving #TODO markers so the user can try to complete the code themselves before Claude provides feedback.

It’s pair programming with an AI that genuinely wants users to grow as a developer.

The concept of vibe coding, which is coding in a collaborative, intuitive, and almost conversational flow, is at the heart of these updates.

With Claude’s new learning features, coding becomes less about producing a quick output and more about enjoying the process of co-creation. The AI doesn’t just “do the work” for users. Instead, it engages in a back-and-forth conversation that feels natural, encouraging users to think, iterate, and refine alongside it.

Anthropic has also trained Claude Code using real-world terminal interactions, giving it a better grasp of developer workflows.

This, paired with the long-context capabilities of models like Claude Opus 4, makes vibe coding more fluid and effective.

Users can have Claude as an AI partner that remembers the flow of their project, understands their intent, and adapts to their working style, all while keeping the collaboration light and productive.

These updates to Claude reflect a deeper philosophy: the most valuable AI isn’t the one that gives the fastest answer, but the one that helps users grow.

By integrating these learning modes, Anthropic positions Claude not only as a productivity tool but as a tutor, or someone (or rather, something) that challenges users to understand, rather than encourage users to just copy and paste the answers and call things done.

In an era where AI can generate code, essays, and business plans instantly, the ability to truly learn from those interactions is a competitive edge.

And with the blend of technical guidance, conversational flow, and personalized mentoring, Claude is becoming a different kind of AI assistant.

Anthropic wants it to make users better at the job along the way.