Purple Hat Desktop is for cozy, manufacturing-style AI development, whilst Fedora Hummingbird is for AI agent experimentation.

Crimson Hat Computer

Of course, Purple Hat has had a computing device distribution for as long as Red Hat has had a Linux distribution. What's distinctive about the AI Developer Edition is that it is primarily based on the Red Hat build of Podman Desktop. Podman is designed to create, control, and install containers on Linux, macOS, and Windows. 

This Linux computing device is constructed on Red Hat hardened pix and Purple Hat trusted libraries to enhance safety. Developers can access those photos and libraries from their laptops while connecting to nearby or faraway OpenShift clusters for unit trying out.

On the OpenShift cluster, Red Hat OpenShift Dev Spaces offers an extensible framework that helps you to combine your preferred AI-driven gear directly into your cloud-primarily based ide. This setup includes a technical preview of the AWS Kiro coding assistant, in conjunction with integrations for Microsoft Copilot, Claude CLI, cline, keep, roo, and extra. Via assisting both proprietary and open-source assistants, Crimson Hat permits you to apply frontier models or not. For this reason, you could code with something that works great for you.

This version of the purple hat laptop additionally presents isolated AI-agent sandboxing through the open-source Kaiden. In turn, this approach allows you to construct and check AI dealers on neighborhood hardware whilst preventing improper AI actions from screwing up your host working device. As a result, you may accurately construct and test AI retailers on your pc.

Pink Hat's superior developer suite additionally provides new skills, along with an AI-powered take advantage of intelligence to modernize your security throughout the software development chain. This feature makes use of AI to determine if the vulnerabilities to your AI-generated code are applicable to a particular application runtime. This functionality offers you the strength to prioritize fixes and remediation based on real hazards.

Fedora hummingbird Linux

Fedora Hummingbird Linux, alternatively, is a free, photo-based, rolling-launch running machine motive-constructed for ai marketers and their developers. The distribution bypasses traditional linux release freezes, turning in upstream updates as quickly as they're available from upstream communities.

In his keynote, Gunnar Hellekson, vice president and worldwide manager of Red Hat Company Linux (rhel), brought, "that is a no-cost, 'loose as in beer and loose as in freedom' running system. If you want assistance for the os, however, Red Hat plans to provide assistance for Fedora Hummingbird Linux included as part of your rhel subscription."

Fedora Hummingbird Linux is hosted in the Fedora undertaking network and supports anonymous, agent-pushed pulls for instantaneous deployment. The distribution removes registration partitions that typically slow AI agent experimentation, aligning with Pink Hat's description of the "on the spot-on expectancies of the agentic era."

This computing device, Linux, is introduced through an agent-stronger, "lights out" ai software manufacturing facility. AI agents perform a good deal of the upkeep and characteristic integration with human-in-the-loop oversight. Built with an equal automated infrastructure as crimson hat hardened photographs, Fedora Hummingbird Linux ships with languages, runtimes, databases, and tools free of known cves and observed via full software bill of materials (sbom).

Understanding the key variations

The 2 offerings serve distinct, complementary roles in Purple Hat's agentic ai strategy: making each service available beneath a single subscription.

Crimson hat plans to make Fedora Hummingbird Linux a default choice throughout developer-focused cloud vendors. Red Hat laptop will serve as the governed, production-mirroring environment extending down to the developer's computer. 

Crimson hat's desire, of path, is that maybe AI builders will start with hummingbird and then move directly to crimson hat laptop and the rest of the pink hat ai own family when they get a job developing production ai applications.

How to Learn Claude Code for Free With Anthropic’s AI Courses – One Course Took Me Just 20 Minutes

If you want to learn how to use Claude Code, AI agents, and Anthropic’s growing ecosystem of AI tools without spending money, Anthropic now offers a surprisingly useful free training library. The company’s online learning platform includes beginner-friendly courses on Claude, Claude Code, MCP, AI agents, and advanced integrations like Amazon Bedrock.

For developers, creators, and AI enthusiasts, these free Claude courses are one of the easiest ways to understand how Anthropic’s AI tools actually work in real-world workflows.

Anthropic has quietly built a free online learning hub filled with video-based AI training. The platform includes short tutorials, in-depth technical lessons, quizzes, and certificates for selected courses.

The available topics include:

• Claude AI basics
• Claude Code tutorials
• AI agents and automation
• MCP (Model Context Protocol)
• Prompt engineering
• Claude integrations
• Amazon Bedrock with Claude

Some lessons take only 15 to 20 minutes, while others are full-length technical training programs lasting several hours.

One of the shortest courses currently available is “AI Capabilities and Limitations,” which takes around 15 minutes to complete. On the other end, advanced learners can explore the eight-hour “Claude with Amazon Bedrock” series featuring dozens of lectures.

My Experience Taking the “Introduction to Subagents” Course

I recently tested the 20-minute “Introduction to Subagents” course because I wanted to better understand how subagents work inside Claude Code.

For anyone unfamiliar, subagents are essentially smaller AI task workers that Claude Code can delegate jobs to automatically. Instead of handling everything in one massive workflow, Claude can assign specialized tasks to separate agents and combine the results later.

Think of subagents like a team of interns:

One handles research

Another writes code

Another checks errors

Another organizes output

This approach helps improve workflow efficiency while reducing overload inside a single AI context window.

The course itself included:

• Four short YouTube videos
• Written explanations
• Examples inside the learning dashboard
• Progress tracking and completion tools

Anthropic uses Skilljar as its learning management system, and registration is completely free.

What I Learned From the Free Claude Course

Even though the course was short, it offered genuinely practical insights.

The most valuable lessons covered:

When to Use Subagents vs Main Agents

The instructor explained when a task should stay inside the main Claude workflow and when it makes more sense to delegate to a subagent.

This distinction is important for:

• Large coding projects
• Multi-step AI tasks
• Research-heavy workflows
• Agent automation systems
• How Subagents Improve Context Management

One major advantage of subagents is better context-window efficiency.

Instead of overwhelming a single Claude session with too much information, separate subagents can process smaller focused tasks independently. This can improve speed, organization, and output quality.

Writing Better Subagent Instructions

The course also covered how to optimize subagent specifications so each one performs a specialized role more effectively.

This includes:

Clear task descriptions

Limited scope assignments

Structured outputs

Defined goals

For developers using Claude Code daily, these small optimizations can significantly improve AI-assisted workflows.

What the Course Didn’t Cover

The course wasn’t perfect.

I would have liked more instruction on:

• Managing subagents while running
• Troubleshooting failed tasks
• Fixing stuck agents
• Monitoring long-running workflows

Still, considering the course is free and only takes about 20 minutes, it delivered solid value and practical takeaways.

Best Place to Start Learning Claude Code

If you are completely new to Anthropic’s ecosystem, the Claude Courses library is the best starting point for structured learning.

However, if you want broader documentation and technical resources, Anthropic also offers its larger learning portal through the:

• API documentation
• Claude developer guides
• Support articles
• Prompt engineering tutorials
• AI workflow examples
• Are Anthropic’s Free AI Courses Worth It?

Yes, especially for developers and AI power users trying to understand modern agent-based workflows.

The biggest advantage is that the courses are:

• Free
• Beginner-friendly
• Short and practical
• Focused on real AI tools
• Updated around Anthropic’s ecosystem

Unlike generic AI tutorials, these lessons are designed specifically around Claude and Claude Code usage.

If you’re exploring AI coding assistants, agent automation, or prompt engineering in 2026, these courses are a strong starting point.

Final Thoughts

Anthropic’s free Claude courses are an underrated resource for anyone trying to learn Claude Code, AI agents, and modern AI workflows without paying for expensive training programs.

Even a quick 20-minute lesson can provide practical techniques you can immediately apply inside Claude Code projects.

For developers building apps, automations, or AI-assisted workflows, spending a little time inside Anthropic’s free training library is probably worth it.

Amazon has confirmed plans to make investments as much as $50 billion in new AI and high-performance computing structures designed especially for US government use.

In a weblog post, the enterprise outlined how the programme will aid a wide range of federal structures that rely upon advanced computation.

This consists of national protection work, scientific tasks, self-reliant structures improvement, cybersecurity analysis, energy programmes, and healthcare-related studies.

Growing reliance on specialized government cloud regions

Amazon claims the funding aligns with federal priorities that call for more potent AI gear deployed on at easy infrastructure constructed in the United States.

The enterprise also links the initiative to broader government efforts to modernise computing ability via upgraded cloud website hosting environments and specialized hardware.

"Our investment in cause-built authorities AI and cloud infrastructure will essentially transform how federal agencies leverage supercomputing,” stated AWS CEO Matt Garman.

“We're giving companies multiplied get admission to to advanced AI abilities so they can enable them to accelerate important missions from cybersecurity to drug discovery. This funding eliminates the era obstacles that have held government back and similarly positions the United States to lead in the AI era."

The investment is expected to add almost 1.3 gigawatts of compute capability across the 3 authorities cloud regions.

Amazon claims this will allow businesses to complete simulation and modeling duties a long way faster than before, with the aid of linking AI systems with real-time feedback loops.

Those projected gains rely on the government cloud garage regions that already assist labeled and unclassified workloads.

Over the last decade, Amazon has constructed GovCloud, pinnacle mystery, and mystery environments to fulfill federal compliance and safety necessities.

It has been a pacesetter in government cloud computing, providing agencies with managed spaces for sensitive computational missions and helping over 11,000 organizations.

This current basis is the platform on which the brand new AI and HPC systems will perform.

It allows companies to undertake expanded compute capacity without rebuilding operational frameworks or shifting away from mounted governance gear.

The funding is part of a wider shift closer to AI-driven scientific and safety workflows.

Federal teams are being pushed in the direction of environments where professional fashions, marketers, and natural-language gear assist research, engineering, and analytical tasks.

Amazon states such skills will let scientists and analysts define complex issues and get hold of automatic hints supported by means of simulations and high-accuracy evaluation.

The enterprise introduced this transition is a break from conventional HPC strategies that require a significant manual effort.