Welcome to the 279th edition of Coding Jag brought to you by TestMu AI!👐
AI is no longer just assisting developers. It’s shaping how code is written, reviewed, tested, and deployed. From AI-generated code and vulnerability triage to autonomous testing agents and natural language automation, the shift is already underway.
This edition looks at what’s real versus hype. You’ll learn what developers actually think about AI coding tools, how many jobs automation may impact by 2030, and how enterprises are preparing for AI agents at scale.
Plus, you’ll find insights on public safety AI, ethical oversight in testing, next-gen QA tools, and events worth bookmarking.
📬 Found something useful or surprising? Hit reply and let’s compare notes.
News
08 min
sonarsource.com
🧑💻 Anirban Chatterjee talks about the latest developer survey by Sonar that cuts through the hype around AI-assisted coding. While developers see clear productivity gains, concerns around code quality, security, and maintainability remain front and center. The report shows where AI helps today and where human judgment is still non-negotiable.
12 minforrester.com
📉 J.P. Gownder at Forrester takes a measured look at the future of work, explaining why automation is more about reshaping roles than wiping them out. Some jobs will vanish, many will change, and tech-heavy functions like QA and operations will feel the shift first.
11 mindevelopers.googleblog.com
🧠 Juyeong Ji at Google DeepMind explains how fine-tuning FunctionGemma turns a base function-calling model into a specialized agent that handles tool selection ambiguity, enforces enterprise logic, and improves performance. He also talks about the FunctionGemma Tuning Lab, a no-code fine-tuning demo for real-world use cases.
09 minperplexity.ai
🚓 Perplexity Team introduces Perplexity for Public Safety Organizations, offering Enterprise Pro to law enforcement with secure, private AI responses that cite sources and work across mobile, voice, and images. It helps agencies analyze reports, transcripts, notes, and multilingual data for better decision-making.
AI
09 minministryoftesting.com
🧪 Matthew Whitaker at Ministry of Testing explains why traditional testing falls short for AI-driven, self-adaptive systems. He highlights the need to evaluate fairness, explainability, resilience, and ethical accountability, and argues that testers must balance autonomous test agents with human oversight and judgment to build trustworthy intelligent agents.
08 minmake.com
💬 Siew Ann Tan at Make explains how natural language is becoming the way to build workflow automations. Instead of wiring modules manually, users can describe what they want in plain English, and Make’s AI (Maia) interprets it, configures the scenario, and tests it automatically.
11 minkore.ai
📊 Deepak Anand explains why AI agents aren’t failing because of technology but because most pilots lack enterprise-ready design, governance, and clear ROI. He highlights what it takes to move from experiments to production-grade agents that deliver measurable business value in 2026.
09 mingithub.blog
🛡️ Man Yue Mo and Peter Stöckli at GitHub Security Lab show how the new Taskflow Agent uses AI taskflows to automate vulnerability triage. By breaking triage into precise LLM-driven steps, teams quickly filter false positives and generate actionable reports with code references.
Automation
11 minthegreenreport.blog
💡 Irfan Mujagic explains how multi-turn RAG systems often lose conversational context, creating subtle failures in pronoun resolution and topic tracking. He outlines practical QA automation patterns, including conversation chains and semantic similarity assertions, to detect context drift and validate coherent, production-ready responses.
08 minqodo.ai
✅ Nnenna Ndukwe at Qodo talks about how AI code generation hits real limits without strong verification. She explains how codified standards, automated enforcement, context‑aware reviews, and outcome‑focused quality gates help teams catch subtle issues and maintain architecture and safety at scale.
Tools
12 minqovery.com
☸️ Mélanie Dallé walks through Kubernetes‑focused CI/CD tools, highlighting how tools like Qovery, Argo CD, GitHub Actions, and GitLab help automate build, deploy, and environment orchestration while reducing manual YAML toil and environment contention.
07 mintestguild.com
🛠️ Joe Colantonio explores low-code/no-code test automation tools that go beyond basic scripting, showing how these tools embed self‑healing, natural language input, and autonomous test capabilities to reduce maintenance and align testing with real development needs.
Video & Podcast
08 mintestingpodcast.com
🎧 🎙️ In Episode 119 of the Testing Podcast, Demi Van Malcot and Simon Tomes highlight how sharing goals publicly sparks motivation and community. Through activities like goal challenges, “Whose goal is that?”, and discussions on long-term development, the episode demonstrates collective support in software quality practices.
12 minwww.youtube.com
🎥 This video by Execute Automation demonstrates how local vision models outperform pixel-based tools in visual testing. By recognizing elements like buttons instead of flagging minor pixel shifts, these models detect meaningful UI changes more reliably, reducing false positives and improving test relevance.
Events
07 minwww.robocon.io
🤖 RoboCon 2026, organized by the Robot Framework community, will take place in Helsinki from February 10-13 and online March 3-6. The event emphasizes interactive workshops, Community Days, and collaborative sessions.