https://iceberg.dbj.org/posts/Recent content in Posts on DBJ IcebergHugoen-usMon, 01 Jun 2026 00:00:00 +0000https://iceberg.dbj.org/posts/post-hype-agents/Mon, 01 Jun 2026 00:00:00 +0000https://iceberg.dbj.org/posts/post-hype-agents/A pragmatic, engineering-first definition of agentic architectures — policy-controlled, deterministic-adjacent task handlers, not autonomous magic.https://iceberg.dbj.org/posts/ea-is-an-instrument/Sun, 31 May 2026 00:00:00 +0000https://iceberg.dbj.org/posts/ea-is-an-instrument/EA does not redesign itself around the thing it is measuring. For 50 years, EA has matured into a vital tool that tells you whether a business is truly aligned with its technology.https://iceberg.dbj.org/posts/feature-not-a-bug/Wed, 27 May 2026 00:00:00 +0000https://iceberg.dbj.org/posts/feature-not-a-bug/LLMs compound prediction errors into hallucinations — and you pay for the noise bundled with the signal.https://iceberg.dbj.org/posts/enterprise-ai-at-last/Wed, 13 May 2026 00:00:00 +0000https://iceberg.dbj.org/posts/enterprise-ai-at-last/Aleph Alpha's on-premises model normalises LLM infrastructure the same way RDBMS was normalised in the 1990s. The procurement model is finally familiar.https://iceberg.dbj.org/posts/language-layer-does-not-matter/Tue, 10 Mar 2026 00:00:00 +0000https://iceberg.dbj.org/posts/language-layer-does-not-matter/On llama.cpp, inference engines, and why the host language is irrelevant when the GPU is the bottleneck.https://iceberg.dbj.org/posts/ai-or-iceberg/Fri, 06 Mar 2026 00:00:00 +0000https://iceberg.dbj.org/posts/ai-or-iceberg/Clear roles, clear communication, safe sailing. How DBJ.METHOD addresses the structural risks AI amplifies.https://iceberg.dbj.org/posts/modular-monolith-for-mainframe/Tue, 03 Mar 2026 00:00:00 +0000https://iceberg.dbj.org/posts/modular-monolith-for-mainframe/Modular monoliths are highly suitable for mainframe modernization. Here is why, and how.https://iceberg.dbj.org/posts/most-valuable-architecture/Fri, 27 Feb 2026 00:00:00 +0000https://iceberg.dbj.org/posts/most-valuable-architecture/CMM → MVA → AI → ROI. Stop funding experiments. Start building equity.https://iceberg.dbj.org/posts/why-cmm-onboarding/Mon, 23 Feb 2026 00:00:00 +0000https://iceberg.dbj.org/posts/why-cmm-onboarding/Transformation fails without measurable organizational readiness. CMM onboarding establishes the foundations before any work begins.https://iceberg.dbj.org/posts/platos-cave/Mon, 16 Feb 2026 00:00:00 +0000https://iceberg.dbj.org/posts/platos-cave/Plato's allegory reframed: the shadows on the wall are your legacy systems, and the light outside is architecture.https://iceberg.dbj.org/posts/incompetence-is-out-of-hand/Thu, 12 Feb 2026 00:00:00 +0000https://iceberg.dbj.org/posts/incompetence-is-out-of-hand/What's actually out of hand isn't AI. It's the organizational incompetence that AI is being asked to hide.https://iceberg.dbj.org/posts/crufty-ai/Sun, 08 Feb 2026 00:00:00 +0000https://iceberg.dbj.org/posts/crufty-ai/You can't get intelligence out of a system that was never coherent to begin with. The fix isn't a better model. It's architecture.https://iceberg.dbj.org/posts/ai-history/Thu, 05 Feb 2026 00:00:00 +0000https://iceberg.dbj.org/posts/ai-history/Kunihiko Fukushima's Neocognitron in 1979, WABOT-1 in 1973, and the deep roots of what is being sold as new.https://iceberg.dbj.org/posts/tonyhoare/Sun, 01 Feb 2026 00:00:00 +0000https://iceberg.dbj.org/posts/tonyhoare/The paper everyone forgot — almost. On Hoare's 1966 record handling, the 35-year mistake, and what Rust got back.