The outcomes of installs and upgrades might be completely different every time, even with the very same mannequin, however it will get loads worse should you improve or swap fashions. If you’re supporting infrastructure for 5, 10, or 20 years, you will be upgrading fashions. It’s laborious to even think about what the world of generative AI will appear to be in 10 years, however I’m certain Gemini three and Claude Opus 4.5 won’t be round then.
The risks of AI brokers improve with complexity
Enterprise “applications” are not single servers. Today they’re constellations of techniques—net entrance ends, software tiers, databases, caches, message brokers, and extra—usually deployed in a number of copies throughout a number of deployment fashions. Even with solely a handful of service sorts and three fundamental footprints (packages on a conventional server, picture‑primarily based hosts, and containers), the mixtures broaden into dozens of permutations earlier than anybody has written a line of enterprise logic. That complexity makes it much more tempting to ask an agent to “just handle it”—and much more harmful when it does.
In cloud‑native outlets, Kubernetes solely amplifies this sample. A “simple” software may span a number of namespaces, deployments, stateful units, ingress controllers, operators, and exterior managed providers, all stitched collectively by YAML and Custom Resource Definitions (CRDs). The solely sane strategy to run that at scale is to deal with the cluster as a declarative system: GitOps, immutable photographs, and YAML saved someplace exterior the cluster, and model managed. In that world, the job of an agentic AI is to not scorching‑patch working pods, nor the Kubernetes YAML; it’s to assist people design and check the manifests, Helm charts, and pipelines that are saved in Git.
