Perspective
The Butterfly in the Machine
August 14, 2026
Why an AI system gives different answers to the same question, what a rounding error in 1961 has to do with your build in 2026, and what it takes to run systems that keep moving.
Industries · Shipyards & Boat Manufacturing
Seventy years of know-how in the shed. One unwritten rule between it and the contracts that keep it full.
The board says three hulls at fit-out and one on survey. The trades list says two short, again. Nine out of ten marine businesses now report the same shortage, and half are already losing revenue to it.
The page you are reading is universal. Pick your region to see the regulators, cases and pressures that reach your operation.
Where are you based?
The record so far
221obligations
tracked across 25 jurisdictions
Every founding document reviewed by the legally trained founder before delivery.
Documents usually within two business days.
Training certificates independently verifiable at southernsky.ai/verify.
Start Here
Read your yard's position in about five minutes. Twenty plain questions in, four readings back: the AI risks as they reach your business, the regulations that already apply to you, the cost of leaving use unmanaged, and the moves that matter most, ranked from the top. Defence and export exposure changes your answers, and the reading accounts for it.
Get your baselineWhere this sits for you
The pressures in your world
Meanwhile the office found its own help: a tender drafted with a chatbot, an estimate sanity-checked against one, a defect photo sent to a tool no one signed off. Each of those got someone home by five. Each one also carried yard information somewhere your defence and export contracts did not contemplate, and those clauses were written as if it could never happen. Two jobs sit in front of most yards now: capture what the senior tradesmen know before retirement takes it, and protect what the contracts say must never leave. Both start in the same place, a written position on how this yard uses AI. Reading yours takes about five minutes.
The trades are short and the knowledge is retiring
Nine out of ten marine businesses report concern about the skilled-labour shortage and 52.1 percent already report direct revenue loss from it. The deeper cost sits behind the headcount: the methods a retiring shipwright carries were never written down, so a departure removes a pair of hands and some of the yard's memory of how the job is done. Capturing that knowledge is the highest-return AI project a heritage yard holds, once the data boundary exists to do it safely.
Defence yards hire from the same pool
Austal alone is seeking about a thousand additional staff, and the wider precinct build-outs target thousands more. Every welder, fabricator and marine electrician a defence prime signs is one a private yard loses or pays more to keep, and the choice bites hardest in a soft quarter.
Refit quotes meet reality when the boat opens
Across the industry, refit projects routinely run 30 to 50 percent over originally quoted scope, driven by what opening the vessel reveals and by thin specification. Variations recover margin and cost client relationships in the same motion. The estimator's spreadsheet is the highest-leverage document in the building, and it is usually one person's.
Documented Work
Documented in the Log Book
AI Governance
70+ year heritage · Government and defence supply chain · 24-month roadmap · Essential Eight baseline assessed
Seventy years of precision vessel construction for government and defence. With limited formal documentation of its processes.
A structured assessment across 5 stakeholders and 8 operational systems produced 6 prioritised opportunities, a phased 24-month roadmap, and a 90-day action plan with defined success measures, including an Essential Eight cybersecurity baseline aligned to government and defence tender requirements.
Read full engagement
AI Training
3 role-based workshop days, designed from a 12-department audit · 18 people certified, from maintenance and marine to finance and the executive team · Every participant left with a working AI setup and a reusable Skill built on a real task from their own role · Weeks later, reported publicly by the client: the team using AI more, sharing wins, and a real shift in the day to day
Eighteen people, ten departments, one shared standard for how AI gets used across the organisation.
A twelve-department audit produced three role-based workshop days that certified eighteen people across ten departments, with the CHART quality method and the data rule embedded across the team and every certificate carrying a public verification page.
Read full engagementThe Opportunity
What AI is already doing
Tender and bid drafting
Staff draft tender responses with a chatbot, under a classification rule that keeps contract-controlled specifications out of public tools and a named reviewer before submission.
Estimating support
An estimator sanity-checks a fixed-price quote and a spec-change impact against AI, with a checkpoint between the tool's output and the priced commitment.
Defect second-opinion
A foreman photographs a defect for a second read from a tool, checked against the surveyor's sign-off.
Procurement email
Supplier chase-ups and procurement correspondence are drafted with AI and reviewed before they go out.
Knowledge capture
The methods a retiring shipwright carries get captured with AI, the highest-return project in a heritage yard, run only once the classification rule names what must not enter an external tool.
WHS and training records
Toolbox-talk notes and training records are drafted with AI and checked before they enter the record.
These are the workflows the prompt library and training stand up, under the standard the classification rule, approved-tools register and human-checkpoint map set, the same artefacts a DISP review or an export file already assumes.
Where to start
Where to start, and where it leads.
Baseline
See what AI could do for your organisation. The AI Baseline Report takes twenty questions and about five minutes, and reads back the use cases returning hours in organisations like yours, the rules that already apply to you, and the moves that matter most, ranked from the top.
Blueprint
For yards with defence or export exposure: risk classification mapped to contract obligations, the classification rule written and briefed, deployment sequencing for knowledge capture and estimating support.
Governance Essentials
For yards without defence exposure: the AI use policy, approved-tools register and training pathway drafted for adoption inside your organisation, with a prompt library that starts the office's first governed workflow, 90 days of keep-current and team education, a recorded briefing, a 30-minute walkthrough call and 30 days of email support. USD $690 founding, then USD $990.
Training day
The office and the shed to one standard, toolbox-talk format, verifiable certificates.
Keep-current
The position holds as defence expectations and EU product liability rules move.
Written from inside your world
Kristina Agustin
Founder & Principal Digital Navigator, Southern Sky AI
20+ years in international superyacht and maritime operations. Legally trained (LLB, Graduate Diploma of Legal Practice). AI educator and consultant. ATSE Elevate Scholar 2026.
Questions
Questions we hear
Yes, inside a written boundary. Defence and government contract data carries handling conditions that one paste into a public tool can breach, whatever the intent. The first artifact we draft is a classification rule: what never enters an external tool, written down and briefed to the office and the shed. Everything else builds on that line.
Start with where you stand.
The AI Baseline Report reads your position in about five minutes, and your answers pre-fill everything that follows.
Run the Baseline
Already know where you stand? Governance Essentials puts your position in writing inside 30 days.