// AI in Practice · Part II · How the Machine Thinks

You Already Know AI –
You Just Called It Search

Khurram Saleem Foundations · Architecture · Agents
Chapter 3 of 18 Primer · 7 min

In 2017, a team at Google published a paper that most developers never read. Eight authors. Nine pages. A title that sounds like philosophy: "Attention Is All You Need." That paper retired the architecture that had dominated natural language processing for a decade – and replaced it with the transformer. If you have used a language model in the last three years, you have been living inside the consequences of that paper.

// the crux

Modern AI is search's descendant. Tokens, attention, vector retrieval, ranking – you already wrestled with these as a search engineer, under older names. Connect the two and the whole stack stops being magic and starts being engineering.

// in one breath
  • Why almost everything in "modern AI" is something a search engineer already wrestled with a decade ago – under an older name.
  • A working search engine taken apart into the very algorithms that, renamed, became AI – autocomplete, edit distance, n-grams – each one you can run yourself, right here.
  • The one move search never made: generating language it had not already indexed. That is the machine itself, and Chapter 4 takes it apart.

But here is the part almost no one says when explaining AI to developers: you already understand the core concepts. You learned them in a different context, under different names. Search engine development – especially the decade between 2005 and 2015 – was applied machine learning at scale. The vocabulary was different. The problems were identical. And if you can connect the two, the entire AI stack stops being magic and starts being engineering.

before & after

The Year the Architecture Changed

Before 2017, machine learning systems processed language sequentially – reading a sentence word by word, carrying a kind of rolling memory forward. The approach worked. It did not scale. Long documents, complex instructions, multi-step reasoning: all degraded predictably as context grew. The model read everything, but attended to almost nothing.

2017
// The inflection point
The transformer replaced sequential processing with parallel processing and introduced attention – a mechanism that learns which parts of an input to weight more heavily when producing an output. The model does not read left to right. It considers all tokens simultaneously and decides what matters.

The result: better language understanding, larger context windows, generalisation across domains. Everything you call "AI" today – GPT, Claude, Gemini, Mistral, Llama, Codex – runs on some variant of this architecture. The transformer is the Intel 8086 of this era: not the final word, but the foundation everything else is built on.

the lineage

Search Was Your First Teacher

The search engine problems solved between 2005 and 2015 were not different problems from the ones AI solves now. They were the same problems, solved with less compute and simpler models. What changed in 2017 was the scale of the solution – not the shape of the problem.

Walk through the concepts side by side. If you built on search infrastructure, or just thought carefully about how search worked, you already have the mental model:

// concept lineage: search → AI
Search (2005–2015) What it solved AI equivalent (2026)
Query tokenisation Split raw text into units the system can operate on LLM tokenisation
TF-IDF / BM25 ranking Score documents by relevance to a query Attention weights
Inverted index Map terms to documents for fast retrieval Vector database / embedding index
Query expansion Retrieve related concepts, not just exact matches RAG / context retrieval
PageRank / link graph Score nodes by their connections, not just their content Graph RAG / knowledge graphs
Personalised ranking Adapt results to user context and history Fine-tuning / RLHF / agent memory
Autocomplete Predict the most likely next token given prior input Next-token prediction (LLM core)
Session context Maintain relevance across a multi-query session Conversation context / in-context memory

If you understand why inverted indexes exist, you understand why vector databases exist. If you understand what query expansion solved, you understand what RAG solves. If you understand PageRank, you understand why GraphRAG retrieves better than flat keyword search for connected data. The problem space did not change. The architecture did. The scale changed dramatically. But the instinct is the same.

Concepts are one thing; the journey is another. Here is what actually happens between a keystroke and a result, taken from a working search engine: two connected engines, Suggest and Result, with a learning loop underneath that quietly gets better every time. Where a stage is a named algorithm you can run, open its try it live panel and step through it yourself, right here on the page.

// the honest version

Most of what gets explained as AI magic is search engineering with better compute and a different name. Start there, and the rest becomes learnable.

That is the search engine, end to end: the lineage from a single keystroke, the two engines that turn it into a result, and the algorithms underneath that you can run for yourself. Every one of them predates the word "AI" by years. The one move search never made was to generate language that was not already indexed somewhere.
// carry forward

You have the lineage: modern AI is search with better compute and a new name. The one move search never made was to generate language it had not already indexed – and that single step is the whole of a language model. Chapter 4 takes it apart the same way.

// read next
AI in Practice · Chapter 4 · 9 min read
Inside the Language Model

You have seen the ancestors. Now the machine itself: the four-layer stack the 2017 architecture produced, and the live AI lifecycle from prompt to answer, with tokenisation, embeddings, attention and next-token generation each one playable, inline.

Read →