Structural Analysis

AI-Native Ways of Working — Software

From software engineering to the liberal arts — the foundational shift of the technical profession

Introduction — why build, and how to start

01

AI Solves the World's Hardest Coding Problems — From the Codeforces 2700 tier to design — in the Mythos/Fable era, AI became the strongest SIer, callable for $200 a month

AI's code-writing ability has matched human top-class on public competitive-programming ratings (the Codeforces 2700 tier). It can also assemble a cyberattack autonomously — evidence of design ability. In the Mythos/Fable era AI became the strongest SIer, handling requirements, design, and build, and callable by anyone for $200 a month. The whole sub-series argues outward from that.

02

Maintenance-Phase Shift Is the Real Story — Cheaper coding is the tip of the iceberg — because AI understands context, maintaining at the code level stops being necessary

The most overlooked consequence of AI writing code is not faster coding; it is the structural shift of the maintenance phase itself. Because AI understands context, maintaining at the code level stops being necessary; the unit of maintenance moves from code to design, spec, and context, and the cost of reading legacy code evaporates. The shift is conditional on keeping design, spec, and context explicit and reconciled with reality; without it, AI-generated technical debt piles up fast.

03

AI Now Does the Software Engineer's Work — The coder obviously goes — and the software engineer who designs and codes becomes, not a designer-coder, but someone who works in dialogue with AI (the builder)

The coder (the code-writing role) obviously disappears. But the subject of this chapter is what lies beyond — AI now does the software engineer's work too. Opus is a coder, Fable/Mythos software engineers; design and code are both carried by AI. What stays with humans is not designing and coding yourself, but building and operating a system in dialogue with AI — that is, the builder. What disappears is the role definition, not people.

04

The Builder Role — Decide what to build, build it in dialogue with AI, run it, integrate the whole

The builder builds and runs whole systems in dialogue with AI — not the next version of the software engineer. The SE solves "narrowly closed problems" and is replaced by AI; the builder handles the "open problem" of raising what to build from reality. This chapter defines the builder as a loop — decide, build with AI, check, integrate — and, along the axis of the SE's narrowly closed problem versus the builder's open problem, shows why judgment cannot be left to AI.

05

Customers Co-Develop with AI — The first move is OSS, not code — use it for the generic, customize it for the personal (ch6–8), build the foundation for the organization (Independence). Only the specific gets written with AI

The era in which customers themselves become builders. But the first move is not writing code — it is using proven OSS. Generic things use OSS first (most economical); personal things are customized on top of OSS in dialogue with AI (the worked examples are ch6–8); organizations stand up a foundation with OSS that replaces Microsoft 365, Copilot, and WordPress (the Independence part). Only the company-specific logic gets written with AI. What AI cannot do, the SIer cannot do either.

Independence — from Microsoft 365, Copilot, WordPress, core systems, and GitHub

01

Becoming Independent from Microsoft and Google — The Whole Map — Move the foundation of your business out of the vendor's cage and into your own hands

The lock-in is not weak features — it is that the layers are closed, and that the key (identity) is concentrated in someone else's hands. The Office suite and the core business systems stand on the same structure. And opening what is closed — that is AI's role. It unseals proprietary formats, reads unreadable code, and extracts imprisoned business knowledge. The Independence part, working with AI, unties the closed bundle into open OSS and moves the key to your side — auth with PocketBase, documents with OnlyOffice, code and sharing with Forgejo, mail with Stalwart, meetings and scheduling with Jitsi/Cal.com, web with Cloudflare Pages, data with PostgreSQL/SQLite/DuckDB, AI with a local LLM + RAG, core logic with FastAPI. This chapter is the map; the chapters that follow stand each one up.

02

Lay the Foundation — SQLite, PostgreSQL, pgvector, DuckDB, Polars — Stand up the data layer everything sits on, first, on your own side

The Independence part starts with the data layer everything sits on. SQLite is usually enough — a single file, no server, already built into Python. Step up to PostgreSQL only when you share and several people write at once. Enable pgvector for semantic search; analyze with columnar DuckDB and Polars — Excel stays the human's I/O while machines crunch the data behind it; pull far ahead of Power BI. The generic is already shared as OSS — you don't write it, you stand it up.

03

Stand Up the Gate — One Login with PocketBase — Share only identity (authentication) — each server enforces its own access control. Central minimal, defense in depth

Authentication is not something each app builds on its own. It is a gate you stand up once and share. But the gate holds only the minimal common thing — identity. Authorization ("what you can do") is held by each app, each server, itself (defense in depth); the single-perimeter idea that "whoever passed the gate goes anywhere inside" is not the design. PocketBase is a single binary with email auth, OAuth2, one-time codes, MFA, an admin UI, and a REST API. Identity lives in the gate; business data lives in PostgreSQL. Step off Microsoft Entra ID's per-seat monthly bill.

04

Bring Code In-House — Forgejo and Zed — The builder's workshop, on your own side — repositories and CI, outside Microsoft

A builder's work is to have AI write code, evaluate it, and integrate it. The place that code lives — the repository — gets stood up on your own side. Forgejo is a single Git forge that replaces GitHub and Azure DevOps, with Actions covering CI/CD. It rides on the PostgreSQL from 2-02 and sits behind the gate from 2-03. The local tools are Zed and Claude. Code is an asset; control of where it lives stays with you.

05

Take Documents Back — OnlyOffice Docs on PocketBase — Office is a format to pass through — take the substance back and embed only the editor engine at the gate

Word, Excel, and PowerPoint are the input/output tools people write and read with — not where the substance lives. Inside Office, AI stays a tool and you stay "the processor." So treat docx, xlsx, and pptx as formats to pass through, not to use, and take back control of the substance. Rather than add a separate storage app like Nextcloud, keep documents as plain files on your own storage, with auth from the 2-03 PocketBase and permissions (access control) in a queryable structured store (PostgreSQL or the gate's PocketBase; xattr when you want zero extra dependency), and embed OnlyOffice Docs as the editor engine only — a thin, custom document app. High-fidelity OOXML means you can still hand anyone a plain .docx. Avoid the finished DocSpace — it brings Active Directory back. The reference implementation is the public repo kura. Keep the formats; take back only control.

06

Mail on Your Own Side — Stalwart and Thunderbird — Outside Exchange and Outlook — receive it yourself, send it honestly

Mail is the record of the business itself. Stalwart is a Rust single server carrying SMTP, IMAP, JMAP, spam defense, and DKIM signing in one, replacing Exchange. It rides on the PostgreSQL from 2-02 and is read with any client like Thunderbird. But outbound deliverability is hard — so lean on an authenticated relay for sending and keep only control of the mailbox on your own side. An honest design.

07

Meetings and Booking on Your Own Side — Jitsi and Cal.com — Teams meetings, Calendly booking, and webinars for classes — on your own domain

Video meetings with Jitsi, booking with Cal.com, classes and webinars with BigBlueButton. Replace Teams, Zoom, Calendly, and Microsoft Bookings with your own domain on your own server. Booking rides on the PostgreSQL from 2-02 and sends confirmations through the 2-06 mail. Step off per-seat, per-minute billing and keep meeting links and records on your own side.

08

Publish the Web — Cloudflare Pages (a WordPress Replacement) — Drop dynamic WordPress and publish a static site with no server — build, verify, deploy as separate steps

Publish, from your own side, the static site you built with AI in the Introduction part. Drop dynamic WordPress and put the baked HTML on Cloudflare Pages — hold no server. Leave the CDN and automatic HTTPS to Cloudflare; keep the source and build in your own hands. The crux is to separate build, verify, and deploy, keeping what you verified identical to what goes live. Internal tools stay behind the gate — borrow the window, hold the vault.

09

Build an API — Expose Core Logic with FastAPI — Rewrite the core system via parallel operation and gather your own logic into one API

'Don't break it, don't touch it' is old advice. AI has cut the cost of rewriting a core system by 10x. Build the new AI-native system in FastAPI, run it in parallel with the old, compare outputs against reality, and when the diffs vanish, kill the old. Push business knowledge out into Markdown all at once, let the floor write the tests, and stop outsourcing. The new logic reads and writes the 2-02 PostgreSQL and verifies identity with the 2-03 gate's token. The reference implementation is kura.

10

Make Your Knowledge Legible — Preparation Is the Main Body, AI the Last Move — OCR, classification, codifying tacit knowledge — move scattered, unwritten knowledge into a written, structured state. A no-regret investment you recover even without AI.

Before you put AI on top, build information worth putting it on. Scattered files, paper and scanned PDFs, tacit knowledge that lives only in someone's head — move them into a written, structured state with OCR, classification, and codification. Preparation is the main body; AI is the last move. What to keep and how to structure it is a judgment only people can make, and it pays off as the end of personnel lock-in even if you never put AI on it — a no-regret investment. Put the prepared information into 2-05's files and 2-02's pgvector, and the next chapter mounts RAG on it.

11

Stand Up Your Own AI — LLM and RAG — Lay AI on top of everything — answers grounded in your own data, on your own side

The Independence part closes by laying AI on top of everything. The pgvector enabled in 2-02 finally pays off. Start by holding Cohere's open-weight coding model North Mini Code locally on Ollama (data never leaves), and load a general model and embeddings alongside for RAG. Embed your documents, code, and mail into pgvector for RAG, and use it through Open WebUI. Keep secrets and always-on processing in-house; borrow a frontier model like Claude for hard judgment — control yours, capability borrowed. Cut the Copilot dependency and close the Independence part.

Shift — the industry consequence

01

Companies Don't Write Their Own Code — Office and Core, Two Parallel Worlds — Writing it in-house was inefficient — so office was bought, core was outsourced, and two worlds stood in parallel

Companies have not written their own code — and that was rational. In-house development was inefficient, demanding a large specialized workforce no single company could justify keeping. So companies BOUGHT packaged software for generic office work (Microsoft) and OUTSOURCED custom core systems (SIer) — two parallel, separately locked-in worlds, joined only by thin seams (auth and document sharing). That parallel split was the efficient equilibrium for decades. AI inverts the efficiency: one person plus AI now stands up both worlds on the same OSS foundation. The premise dissolves, and both vendor structures collapse together. This chapter sets the premise for the Shift part.

02

Digital Sovereignty — The Microsoft Problem and the Trump Problem — OSS and sovereign AI are now the better choice on both economics and security

Until recently, Microsoft 365 was the economical and safe default — which is exactly why everyone bought it. That premise has inverted. OSS plus sovereign (self-hosted, local-weight) AI is now better on both cost AND security. The Microsoft problem — per-seat rent that only rises, data on a US company's cloud reachable under the CLOUD Act, opaque telemetry, Copilot routing content through Microsoft's models. The convenience IS the dependence. The Trump problem — depending on US Big Tech means depending on the US government's goodwill, and the Trump administration cannot be trusted not to weaponize that dependence (sanctions, cut-offs). So leaving Microsoft is no longer ideology — it is the new economic-and-security rational default. This chapter sets the premise for the Shift part's office (Microsoft) side.

03

The Structural Uneconomy of the SIer Model — The upstream judgment outsourcing cannot remove, and the de-responsibilization and hollowing-out it brings — for the same effort, you can build it yourself

Even when you outsource, the upstream judgment — improving the business operations themselves, and understanding the systems — stays with the customer. That work is not a straight line but a loop, and with AI you can run it fast in-house. And the deepest problem with commissioning is the erosion of responsibility and capability: the moment you hand it off, no one owns the result whole. For the same effort, you can build it yourself.

04

The Lock-In Problem — Proprietary frameworks, proprietary abstractions, human dependency — with Palantir's FDE as the archetype

SIer commissioning anchors customers with three layers of lock-in (proprietary frameworks, proprietary abstractions / Ontology, human dependency). Palantir's FDE model is the extreme form — maximizing all three to sustain premium pricing in the tens of billions of yen. AI-native development, by contrast, structurally avoids lock-in: AI tends to write in standard libraries and standard formats, so another AI, another builder, or the customer themselves can take over.

05

Companies Hire Builders — The senior builder is management — moving into the CIO's seat, with heavier responsibility than today

In the AI-native era, the professional work — including lawyer- and doctor-level judgment — is done by AI. So the human senior builder is not a profession that sells judgment but management that makes business decisions: the CIO (Chief Information Officer). IT judgment = business judgment = management judgment, and because IT becomes the core of the business rather than a surface layer, the responsibility is heavier than today's CIO. A general-employee grade cannot accommodate this. The corporate-website case shows both the cost and the structural change.

06

Japan's SIer Industry Transition and Labor Mobility — Multi-tier subcontracting, paradoxically, makes the transition easier

Japan's multi-tier subcontracting structure in the SIer industry is usually treated as a barrier to transition. Dissect the structure and the conclusion reverses — because coder demand is externalized through contracts, the structure can shrink without internal lay-offs. Prime contractors can downsize by not renewing subcontractor agreements; talented subcontractor coders flow to primes, customer companies, or independence. Labor mobility is trending upward, with long-term commissions, secondment, and internal ventures absorbing the shift as transitional forms.

07

The Structural Transition That Won't Reverse — The chain of change, a near-term horizon, and irreversibility — because the premise has inverted, it does not move back. But bounded to software development

From AI reaching top-tier capability through coder displacement, builder demand, and SIer shrinkage, the changes chain together and the main part happens in the near future. And because the premises of economics and security have already inverted, the structure that has moved does not move back. But this "complete replacement" applies only to verifiable-correctness domains like software; in desk work, self-driving, and robotics, AI is blocked at the last 1% and complete replacement does not happen — those are productivity-gain stories. The writing of this very sub-series is itself evidence of that bound.

Align your tools with the AI era, and you become its free person.
The time freed flows into culture, science, and reality.

Start with Chapter 1

AI Solves the World's Hardest Coding Problems