form from.Direction 02 / Research & blueprintCurrent site ↗

07 September 2026 · A proposal for the next build

A lab you can enter.
Work you can inspect.

Keep the formation. Bring back the cellular field. Give the engineering real depth—and each part of the site a different way to be experienced.

01 Encounter02 Inspect03 Explore04 Verify05 Read06 Meet

What the visual walk found

The same frame.
Different words.

The current page repeats a small label, a large heading, two columns and a divider. Changes in background don’t change the reading experience.

The next version needs an artefact, a living surface, an architectural drawing, an editorial spread and a human introduction. Their content should determine their composition.

The proposed homepage

Six encounters.
One considered sequence.

Composition studies, not finished page designs. Select a chapter to inspect its role. The field reuses the existing live experiment.

01 / A sculptural opening

Give the possible
a form.

The existing chunky graphite formation.

Applied R&D.
Integrated AI systems.
Built to operate in Australia.

A direct route to Systems ↗

A clear offer, before the first scroll.

Retain the object and transformation. Shorten the extended scroll stage. Put Systems and contact within immediate reach; make the combination thesis a brief transition.

Form
One object, generous space
Motion
Scroll-driven compression
Keep
Current geometry and film
02 / System dossier

Operating
knowledge.

Proposed internal reference build

Constraint → Decision → Evidence
Illustrated scenario01 / 03

Which procedure
applies?

Procedure · Revision 02Superseded
Procedure · Revision 03Current

The answer should carry its source. The action should wait for its owner.

Action held for review
A legible artefact leads the story

Show the difficult engineering decision.

Open with a source and result, then let the reader inspect a conflict. The dossier explains retrieval, integration, evaluation and human control. A real build must supply the evidence.

Form
Large artefact, margin annotations
Interaction
Inspect a consequential choice
Status
Scenario until implemented
03 / A study in emergence

Small rules.
Unexpected worlds.

Disturb the field.
Observe what follows.

Branching monochrome contours from the existing reaction-diffusion simulation.
Material study

A permanent part of the lab’s identity.

The field returns to the homepage and gets its own Lab page. Prototype fewer seeds and larger structures next. This preview preserves the existing simulation; it is Gray–Scott reaction–diffusion.

Form
A continuous dark surface
Interaction
Disturb, observe, reset
04 / Reference architecture · Proposed

Where it runs is
only the first question.

Australian operating boundaryAccess & responsibility
Your sourcesData & permissions
InferenceModels & routing
A decisionHuman authority
KeysLogs & recoveryDependencies & exit

Who can access it. Who holds the keys.
What leaves the boundary.

Make the operating boundary understandable.

A quiet architectural drawing follows one request and identifies the responsibilities around it. Keep the essential facts visible. Actual providers and access arrangements still need to be specified.

Form
Wide section drawing
Pace
Still, precise, readable
Depth
Working together → Control
05 / From the labProposed editorial feature

The economics
of useful
intelligence.

The cost of a useful answer includes every attempt that failed to produce it.

Routing, retrieval, verification and human review all belong in the calculation.

A model call is one part of the task.
ModelComplete task →

Let the reader encounter the thinking.

Feature one substantial excerpt, a useful margin figure and a confirmed byline. Give articles their own URLs. Link the economics essay to the cost–quality instrument in the Lab.

Form
An asymmetric editorial spread
Typography
Serif as a reading voice
Keep
The strongest original writing
06 / An open invitation

Bring us
the hard part.

Real portrait
to be supplied

A conversation with
the people doing the work.

Confirmed technical lead
Relevant experience · Contact
Start a conversation ↗

Give the lab a human presence.

An authentic portrait and a specific introduction to the technical lead. Explain the first conversation: the capability sought, the current system and the constraints. Real names and details come before publication.

Form
A personal introduction
Evidence
Confirmed experience
Next step
A clear conversation route

The proposed sitemap

A selective homepage.
Depth behind it.

Four primary destinations and a conversation route. Build depth around the work we actually have; add records as evidence develops.

HomeSix encounters → relevant depth
01

Systems

Evidence register

  • Operating knowledge Proposed internal reference
  • More systems as work exists
02

Lab

Experiments & writing

  • Emergence Existing cellular field
  • Cost & quality Existing synthetic study
  • The economics of useful intelligence First editorial feature
03

Working
together

A programme brief

  • Question, baseline & constraints
  • Evidence & decision gates
  • Australian operating boundary
  • Ownership & handover
04

About

The people & position

  • Confirmed technical leads
  • Relevant experience
  • Why Form From exists
ContactThe problem. The constraints. A real person to speak with.

These are planned destinations. The storyboard and the existing studies are the only new-direction previews available now.

Research → design decisions

Borrow the lesson.
Develop our own language.

Primary sites, engineering case studies and scientific publications. These are qualitative observations, not evidence of conversion performance.

01 / Experimental practice

Give curiosity a permanent address.

Lusion maintains a separate Labs destination and describes self-initiated work as part of its practice. Form From’s cellular experiment deserves an equally deliberate role.

Lusion’s first-person studio account ↗
02 / Interactive explanation

Let a reader test the idea.

Distill puts a manipulable model beside an explanation of its behaviour. The field and the cost study should expose cause and consequence, with room to investigate.

Distill · Growing Neural Cellular Automata ↗
03 / Applied R&D evidence

Show the decision that changed the work.

TTP’s ACT Medical record includes prototypes, an unselected concept, testing material and named contributors. That evidence structure is useful for an integrated-system dossier.

TTP · ACT Medical case study ↗
04 / Research publishing

Let substantial work set the pace.

Physical Intelligence’s quiet index leads to articles with films, figures and papers. Form From can use a restrained shell while letting each artefact determine its presentation.

Physical Intelligence · π0.5 article ↗
05 / Commercial credibility

Make the status of the evidence clear.

Faculty’s NESO case identifies its proof-of-concept scope and qualifies potential benefits. Our proposed build, synthetic study and future client results need equally accurate descriptions.

Faculty · NESO VoltaVisor ↗
06 / Australian control

Explain who operates the system.

Australian guidance addresses accountability, testing and supply-chain responsibilities. Our control story should identify access, keys, operations and dependencies within the actual stack.

National AI Centre · Implementation guidance ↗
Other references and scope of review

Sakana AI — a spare identity with a coherent natural motif. Thinking Machines Lab — a considered editorial position. Hosting Certification Framework — ownership, control and supply chains within its government hosting scope. None implies that Form From is certified.

Accessed 7 September 2026. Desktop review sampled openings and lower sections; it was not a complete interaction or device audit. Some Faculty media remained unloaded. Active Theory returned an unsupported-browser screen. Lusion’s opening rendered on a follow-up attempt; its deeper scenes and Labs experience were not fully reviewed. Its navigation and first-person account support the structural findings. No third-party artwork has been reused in this storyboard.

The build plan

Resolve the whole.
Then refine the details.

  1. Complete

    Research & blueprint

    Visual audit, sources, sitemap, the six encounters and evidence requirements.

  2. Next

    Composition prototype

    A separate Home and System dossier. Review desktop and mobile together before polishing motion.

  3. Evidence

    Make the work inspectable

    Build or document the reference system. Gather confirmed people, actual stack details and measured results.

  4. Implementation

    Build the deeper pages

    Systems, Lab, working together, About and contact. Reuse the existing formation and simulations.

  5. Finish

    Motion, editorial & validation

    Review the complete journey, responsive compositions, keyboard use, reduced motion, fallbacks and publication claims.

The material design cannot invent.

One documented system. Confirmed technical leads. The actual Australian stack and operating arrangements. Real contact details. We can develop the composition with labelled placeholders while those facts are gathered.