
Jev and the Case for a Model That Only Decides
Some conditions are easy to write in software. If a basket contains three or more items, offer free delivery. Others are much harder.
Subject
What happens inside AI: how models answer, reason and retrieve, and where their mistakes come from.

Some conditions are easy to write in software. If a basket contains three or more items, offer free delivery. Others are much harder.

I was recently asking myself a really obvious but strangely complex question: if all a large language model does is predict the next word, how can it actually reason?

There's been a lot of talk about "context engineering" in 2026 i.e. giving AI access to documents and information. But what does it mean when an AI system "finds a document"?
These articles explain how AI actually works, step by step: how a language model produces an answer and appears to reason, why the retriever matters as much as the model, where confident mistakes come from, and the chips that make it all possible. Start here if you want to understand the technology before deciding what to do with it.
| No. | Title | Published | Reading time |
|---|---|---|---|
| 138 | How Do We Reach Continuous Learning? | 10 min | |
| 137 | Jev and the Case for a Model That Only Decides | 9 min | |
| 130 | Does Context Engine replace the CMDB? | 5 min | |
| 128 | If AI just predicts the next word, how does it think? | 11 min | |
| 127 | The Search Retriever is not the LLM | 9 min |
| No. | Title | Published | Reading time |
|---|---|---|---|
| 120 | Is AI Confidence Your Biggest Enterprise Risk? | 5 min | |
| 117 | From Pixels to Powering the Future, How Nvidia Built the Brains of Modern AI | 10 min | |
| 109 | RAG in ServiceNow | 4 min |
Decide the pattern
AI Solution Decision Framework (opens tools.thedigitaliceberg.com)Guide teams toward the right AI solution pattern with a practical, step-by-step assessment flow.
Issue 04 June 2025
Creating the Optimal AI Solution ArchitectureA practical, architecture-led guide to designing, scaling, and governing AI solutions using structured frameworks, strategic alignment, and enterprise platform thinking.