#AI horizons 25-11 – AI Agents Can’t Actually Do Your Job (Yet)
Analysis of AI agents' current limitations, showing they complete only 2-3% of real freelance tasks, highlighting the gap between automation hype and reality.
Analysis of AI agents' current limitations, showing they complete only 2-3% of real freelance tasks, highlighting the gap between automation hype and reality.
Developer shares custom Claude Code plugins 'Essentials' and 'Ideation' that use AI to clean code and structure ideas while preserving personal style and judgment.
A technical guide on building an AI agent server using the AG-UI protocol and Microsoft's Agent Framework for Python, with GitHub code.
A developer uses the Goose AI agent and Claude Sonnet 4.5 to build and deploy a themed winter festival website to Netlify, detailing the technical process.
Microsoft's AI strategy pivots from exclusive OpenAI partnership to a multi-model platform, focusing on governance, enterprise integration, and competing in the agent and edge AI markets.
A developer uses the Goose AI agent to build data visualizations for an AI coding challenge, documenting the process and troubleshooting tool issues.
Explores advanced Context Engineering techniques for AI agents, focusing on combating Context Rot and improving multi-agent coordination.
A developer builds a GitHub Action using the Goose AI agent to generate daily fortunes with ASCII art and commit them automatically.
A developer-friendly introduction to the Microsoft Agent Framework, an open-source SDK for building and orchestrating AI agents in C# and Python.
Explores 'context plumbing' for AI agents, the engineering needed to move relevant context from various sources to where AI systems run.
Brendan Gregg discusses AI agents trained on his performance engineering work, their limitations, and the ethical implications of creating 'Virtual Brendans'.
Senior engineers struggle with AI agent development due to ingrained deterministic habits, contrasting with the probabilistic nature of agent engineering.
A monthly tech link roundup covering AI agents, Kafka, Flink, LLMs, conference tips, and commentary on tech publishing trends.
A step-by-step tutorial on building a functional AI agent using the Gemini 3 Pro model and Python, covering core concepts like tools, loops, and context.
Discussion on Kubernetes' suitability for AI workloads and Google's new AI agent technology, following KubeCon 2025.
Guide to using DevUI and Microsoft Agent Framework for debugging and visualizing AI agents in .NET applications.
Experiment testing if AI vision models improve SVG drawings of a pelican on a bicycle through iterative, agentic feedback loops.
Weekly summary of AI industry trends covering enterprise adoption, chip manufacturing, regulation, robotics, and major company announcements.
The article argues that writing a simple AI agent is the new 'hello world' for AI engineering and a surprisingly educational experience.
Moonshot AI's Kimi K2 Thinking is a 1 trillion parameter open-weight model optimized for multi-step reasoning and long-running tool calls.