Becoming Engineers

Barnaby Robson & Chau Woeste
September 2026

Becoming Engineers.

Part one

A new approach to knowledge work

Barnaby Robson

Part two

How do we help people make this journey?

Chau Woeste

Part three

Practical agentic engineering

Barnaby Robson

9 months, 40 billion tokens later

hover to trace · click to open · drag to spin

Terminal

barnaby@Mac ~ %

A file we
can both read.

project-notes.md

# Project notes

## The idea
Make the work easier to repeat.

## Next steps
- [ ] Try one example
- [ ] Write down what worked

Make it agent-native.

Four swaps that make a file agent-nativeWordMarkdownPlain text. Both of us read it.PowerPointHTMLA page. Both of us see it.ExcelCSV + PythonTables the code can check.FoldersYAML properties + JSONThe file says what it is.

What the agent reads well · the formats this talk was built in

File over
Folder

Project Cedar - investment thesis.md

---
deal: Project Cedar
sector: Business services
geography: Hong Kong
workstream: Commercial DD
status: In progress
owner: Barnaby
---

# Investment thesis
Revenue depends on repeat customers.

Illustrative deal note · YAML properties above Markdown

WikiLinks.

Properties say what a file is.
Links say what it touches.

Illustrative notes · fictional deal

“I know
Kung Fu.”

Neo opens his eyes after downloading martial arts skills in The Matrix.

The Matrix, 1999

A candy shop
for adults.

GitHub. Skills, hooks and agent workflows
from the frontier. Open source. Free.

Four adults laughing in a sweet shop, filling bags from jars under a Free Candy sign

Supplied photograph

Automation tools.

Cron job

Runs at an exact time,
again and again. It can trigger
an agent: a morning briefing.

Hook

Reacts to a trigger.
Something happens, and the
agent takes an action.

Remote agent

Always on. An agent that
lives in a server and works
while you sleep.

Cron jobs, hooks and a remote agent · the automation that outlives the session

YOLO.

--dangerously-skip-permissions

Let it get on
with the job.

Set the blast radius first.

Nike Just Do It poster showing people jumping from a cliff into water

10 billion tokens
later…

Context engineering.

Context engineering: capture, stamp, link, answerData roomDatasite exportManagement callsTranscriptsEmailDeal inboxWebWeb ClipperCron · 02:00Hook · on saveRemote agent · overnightCedar vaultvdr-customer-contracts.mdmarket-report.mdvdr-monthly-accounts.mdcfo-email.mdmanagement-call.mdvdr-board-minutes.mdPROMPTCustomer concentrationvdr-customer-contracts.mdRevenue cut-offvdr-monthly-accounts.mdKey-person dependencymanagement-call.mdMarket share erosionmarket-report.mdWorking capital buildcfo-email.mdRaw. Markdown. No properties yet.Properties stamped. Passport checked.Linked. Filed. Fresh when you open the laptop.Five linked notes read. The source is on every line.

Illustrative pipeline · data room, transcripts, email and Web Clipper to Markdown, cron, hooks, a remote agent · fictional Project Cedar

I had nailed
context engineering.

Rocky, the rock-like alien of Project Hail Mary, arms raised against a starfield, with the word AMAZE repeated three times above

Supplied image · Rocky, Project Hail Mary

I am AI guru

Starting out“AI guru”

Illustrative confidence · learning points in story order

Supplied Big Lebowski image framed as TIME Man of the Year

403

“OAuth authentication
is currently not allowed
for this organization”

Recorded error · 5 April 2026 UTC · reconstructed display

Multi-model.

Keep the work when
the model changes.

Test the fallback before you need it.

Supplied black-and-white group photograph, used as a visual pun on multiple models

Supplied photograph · visual analogy

Margin Call API Call.

Monitoring token use and spend becomes a core competency.

Audit where
the tokens go.

Forgotten scheduled jobs took most of my allowance.

Orchestrate
and delegate.

The smart agent plans. The cheapest capable model does the work.

Avoid API
access.

A seat spends an allowance you already hold. Metered credit just goes.

Supplied still: the overnight meeting scene from the film Margin Call

Supplied still · Margin Call (2011) · the bill I could not explain: image generation routed to a metered API.

Less is more.

The model could not see what
68 of my 127 skills were for.

Skill-listing diagnostic · 12 May 2026

A cathedral.
In one prompt.

Sagrada Família surrounded by construction cranes

The cathedral plan, as a visual metaphor.

Agree what good looks like.

The brief

What are we trying
to achieve?

One slice

Show me one complete
example.

The check

Does this do the job
we agreed?

The approved Neo slide from the first slice of this talk

We started this talk
with a slice.

The approved first slice of this presentation · September 2026

Deterministic v stochastic.

Scripts set the bounds. The model chooses what goes inside.

A mock slide: fixed geometry contains a model-selected chart and commentaryDeterministic · fixed by scriptsStochastic · chosen by the modelQuarterly reviewLOGOCommentaryselected bythe model.Font sizeLogo locationText and chart positionsFixed character limitChart typeCommentaryThe model writes insidethe fixed text box.

Did the change improve the work?

A/B test · the same job, with and without the change

Priority-tweets jobBaselineRound 1Round 2
Token budgetPassPassPass
QualityFailFailPass
PresentationFailPassPass

Diff · read what changed before you accept it.

Recorded comparison · 25 June 2026 · pass/fail against my own rubric

Blast radius: how far a mistake travels.

Before you give an agent a new power, know how far its worst mistake can travel. Then shrink it.

Four rings of blast radius, two reported incidents and four controlsAn agent in productionradius: your customersA server with credentialsradius: your accountsAn agent in your filesradius: your vaultA chat answerradius: you13 hoursof outage after a coding agent deleted live AWS production.9 secondsfor an agent to wipe a database and its backups.AllowlistsConfirmation gatesNarrow scopesCrawl–walk–run

PC Gamer on AWS/Kiro; Amazon attributes the outage to misconfigured access controls. Tom’s Hardware on Cursor/PocketOS.

Security.

The agent guardrails say no to five things.

Wipe the disk.
Delete the home folder.
Force over shared history.
Make the private repo public.
Dump the passwords.

A film still of a marshal standing in a doorway, arms folded, blocking the way

brun-guard, 14 September 2026 · after David Ondrej’s agent guardrails · still supplied

The review seat sits with you.

2,797

Source total

2.80

Rendered total

The chart looked right.
Open the file. Check the number.

Recorded totals · 9 June 2026

“Is this good?”

“Excellent.
Ready to share.”

The taste seat stays with me.

Supplied wardrobe photograph: clothes on a rail, chosen by eye

Illustrative exchange · the judgement lesson comes from the May essay · image supplied

Show me.

The SPA

cedar-spa.pdf · 124 pages7.3 Within 60 Business Days after the Completion Date, the Purchaser shall prepare and deliver to the Sellers a draft of the Completion Accounts, prepared in accordance with the Accounting Principles set out in Schedule 5 and, to the extent not inconsistent therewith, the Specific Policies. 7.4 The Sellers shall have 30 Business Days following receipt of the draft Completion Accounts to review them and may, before the expiry of that period, deliver a Dispute Notice to the Purchaser. 7.5 If the Sellers do not deliver a Dispute Notice within that period, the draft Completion Accounts shall be final and binding. 7.7 Any Disputed Item not resolved within 20 Business Days shall be referred to the Independent Accountant…
“Show me the price mechanism as a timeline. Every date, and what happens next.”

See it. Then challenge it.

cedar-price-mechanism.html

Price mechanism: completion accounts

Project Cedar · SPA, 124 pages · click a date

Illustrative HTML view · fictional SPA · Project Cedar is invented

Neo surrounded by Agent Smith copies in The Matrix Reloaded

Cognitive overload.

The Matrix Reloaded, 2003

I had built
myself
a second job.

A sysadmin. For myself. Unpaid.

Illustrative portrait supplied by Barnaby: serving a takeaway bag at McDonald’s

40 billion
tokens later…

Voice.

You can talk 3x faster
than you type.

Speak, don’t type.

Supplied still: a man talking to his agent from bed, from the film Her

Supplied still · Her (2013), as a visual analogy

Rewire your brain.

How do we always deploy new technologies?

Absorb

Doing the old things better with the new technology

Innovate?

Things that are only possible with the new technology

Disrupt?

Redefine the question

Frame and wording after Benedict Evans, “What’s this for?”, June 2026.

Systems thinking.

The Apollo 17 Blue Marble: one connected wholeThe skill underneath it all.A system is the pattern of how its partsaffect each other over time. Change oneand the others react. Miss a connectionand the pattern breaks.Senior engineers built this senseover years of failure.With agents, you need it on day one.NASA, Apollo 17, AS17-148-22727, 7 December 1972.

After the framing in Machinations,
“Is this the only skill left?” (April 2026).

An investment example of systems thinking.

My own research system.

My research system end to end: capture, organise, process, execute, and the loop backCAPTUREORGANISEPROCESSEXECUTE→→→Fiscal.ai / Capital IQ · fundamentalsFMP · pricesFRED · macroPolymarket · oddsX · postslast30days · Reddit, HN, YouTubeResearch vaultequitiescryptocommodscashsignal ledgerwatchlistanalystsenior analystquantIC · DruckenmillerIC · Munger, MauboussinIC · Koch, Perrunaportfolio managerJunior analyst tear sheetSenior analyst reportIC memoFull dossierCron, every morning. No model involved.Every file stamped and linked. One ledger.Analysts first. Then five investors vote.One page, then the full dossier.Every decision re-enters as information.VISTan oil producerA signal in the feed.Filed under equities, linked to the oil thesis.Screened, deep-dived on the fiscal.ai databook, base-rated, voted, sized.Tear sheet, IC memo, dossier.

The author’s own research system · sources as named · one holding shown, no position data

Building in teams.

Everyone pulls from the source of truth. Anyone can suggest a change by branch.

Everyone pulls from a shared repository; anyone can propose a branch for review and mergeEveryoneTeam memberTeam memberTeam memberShared GitHub repositorySource of truthmainPULLEveryone starts here.BranchSuggest a changeReviewCheck the changeANYONEMergeafter review

When ideas
have sex.

Matt Ridley. His name for how innovation actually happens.

Do not productise.
A black box cannot be forked.
Your edge is the rate the file improves.

Matt Ridley on a TED stage, a projected network image behind him

Matt Ridley, “When ideas have sex”, TED 2010, and The Rational Optimist (2010) · photograph supplied

Still learning.

My confidence: rise, collapse and a modest recovery, with the lesson at each stepHow much I thought I knewCommand lineAgent-native formatsEvery file's passportConnected knowledgeSkillsAutomationYOLOCapture, stamp, linkContext engineeringAI guruTest the fallbackToken efficiencyPrune the contextThin slicesDeterministic v stochasticA/B and evalsBlast radiusSecurityHuman judgementShow meCognitive overloadAI idiotSpeak, don’t typeRewire your brainSystems thinkingBuilding in teamsStill learningAnyone whose work passes through a digital interface, which means everyone, will have to knowwhat a skill is, what a hook does, what context engineering means, and why blast radius matters.We are all going to have to become engineers.Confidence is illustrative; spacing follows the story. Grey labels were learnt the hard way.

You wake up. Friday this time. The server is quiet.

I am one person doing this
and finding it hard.

The corporations that have not started are running out of time to find it hard cheaply.

We are all going to have to become engineers · barnabyrobson.org, May 2026

How do we help people
make this journey?

Chau Woeste

Partner, People in M&A · KPMG UK

AI can produce
the work.

Leadership decides
whether it is worth anything.

Output is cheap now.
Craft is what people still pay for.

A shoemaker finishes a leather shoe beside a production line.

AI's workforce collision.

The panels compare job seniority with age groups from a separate dataset.Change in postings requiring AI skills, by roleUS and UK, January 2024 to January 2026Entry levelManagersExecutives0%Reported AI use at work, by ageFigures reproduced from KPMG’s comparison chartAges 18–34Ages 35–54Ages 55–74-12%+71%+472%47%30%23%The comparison suggests a gap between hiring priorities and AI use.

KPMG, AI's workforce collision, 2026: Revelio AI-skill postings and embedded comparison chart. Age-use figures are reproduced as published; their cited ECB source differs. kpmg.com/xx/en/our-insights/ai-and-technology/ai-workforce-collision.html

AI changes how people learn to think.

MIT reports that AI can give credible answers to almost any written assignment inits undergraduate curriculum.MIT undergraduates surveyed, autumn 202590%were concerned aboutoverreliance on AIlanguage models67% were very concernedWhat the MIT committee recommendsReconsider what written work proves.Put oral, social and hands-on work back intoassessment.State the permitted uses of AI for each course.How do people build judgement when AI supplies the answer?

MIT Ad Hoc Committee on AI Use in Teaching, Learning and Research Training, 2026. Autumn 2025 MIT Tech survey, n=659 undergraduates. aiandeducation.mit.edu/report; /appendices, C.3.

White-collar work
is entering
mass production.

Machines make more,
faster, for less.

Craft, judgement and reputation
are what is left.

An experienced shoemaker guides an apprentice inspecting a leather shoe.

A founder will not sell
because the model
is elegant.

They have to believe you know
what they built.

What they are afraid of losing,
and what happens to it next.

The founder looks across the shoe workshop after the shift.

Data is the beginning of the story.

KNOWFEELDO

The AI-enabled M&A leader.

AI expands task capacity within the leader’s role.M&ALEADERSHIPTASKSGet the work donefaster and betterKNOWLEDGE+ JUDGEMENTKnow what mattersChallenge the answerFind the edgeRELATIONSHIPSBuild trustand move people82%say future C-suite leadersneed a strong understandingof generative AI41%are confident they have theskills to implement AIJudgement and relationships help people act on the work.

Russell Reynolds Associates, H1 2025 Global Leadership Monitor. russellreynolds.com/en/insights/articles/leaders-views-on-generative-ai-in-2025

Questions

Barnaby Robson & Chau Woeste

← → to navigate
≡