01 / Journal
Journal
How AI changes software development and IT operations. New roles, narrowly scoped permissions and the responsibility behind agentic systems.
AI news
New AI products, briefly assessed.
What new tools promise, what the sources support and which questions remain open.
- 01Jev: TypeSafe's System One decision modelJev is not a coding agent, but TypeSafe AI's System One model. Typed probabilities instead of chat — and what is public about cost and usefulness.↗
- 02Grok Bot hype check: What the AI agent can really doGrok Bot promises always-on AI agents with a cloud computer. What is genuinely new, where the hype holds up, and which limits businesses should examine.↗
Practice & analysis
More posts
Explore ideas in depth, connect the dots and consider how technology fits into everyday life.
- 01How AI is changing DevOps in 2026.The newest shift is from AI assistance to bounded agents across delivery and operations—with review, telemetry and platform guardrails becoming more important.↗
- 02Agents, permissions and responsibility.Operational risks are easier to understand through access and decisions than the image of a malicious agent. Where effective boundaries belong.↗
- 03Agents in operations: permissions by task.Integrating AI into DevOps and system administration through separate identities, scoped tools and clear boundaries between reading and changing systems.↗
- 04Promoted to pull request reviewer?When agents write code, the work shifts toward defining tasks, reviewing changes and taking responsibility for the result.↗
- 05Chat Control: The Majority Voted No — and Lost Anyway314 MEPs voted against Chat Control, 276 for it — the law is in force anyway. The Council kept Parliament's encryption carve-out.↗
- 06Welcome to MainframeWhat Mainframe stands for and which practical guides to automation, AI, infrastructure, and security will appear here.↗