Skip to content
● Index of systems, updated 30 Sept 2026

Systems: built, modelled, argued

Lorenzo Roque
Dal Fabbro

I work on systems. I model economies, organisms and games; I build the software that runs my own life; and I write about the institutions that run everyone else’s.

Black-and-white portrait of Lorenzo Roque Dal Fabbro

Politics & Philosophy, University of Sussex, class of 2027.
Born in Lisbon in 2004; writes in English and Portuguese.

Running now

  • PolyVision · training PPO agents on 5,517 real Polytopia maps
  • Econ-Engine · MVP8: a 100-household economy with wealth-tax experiments
  • Kleos · methodology 2.0.1 for evidence-based self-assessment
Live systemn=480 · seed 2004 · t=00000

Rules

Fig. 0 Particle life: three species and nine numbers. Every cluster, chase and orbit emerges from the rules. Move across it to perturb the system; edit a rule and watch it reorganise.

§ 01System map

Everything connects.

Five subsystems on one timeline. Lines are lineages, stations are projects, dashed transfers are arguments that later became software. Select anything on the map.

Scroll sideways to follow the lines →

May2025JunJulAugSeptOctNovDecJan2026FebMarAprMayJunJulAugSeptOctFLAGSHIPPROJECTESSAYCOLUMNARGUMENT → SYSTEMACTIVENOWS1Economies3 PROJECTSS2Strategy & worlds5 PROJECTSS3Artificial life8 PROJECTSS4Personal systems5 PROJECTSS5Institutions41 PIECES34 essays and 7 Observador columns in eight monthsEcon-EngineUn-Dead HotelPolyVisionAtlas CoreParticle SimulatorParticle LifeLife Engine 3DAriadneHeraclesKleosActivityGITHUB CONTRIBUTIONS / MONTH2025-05: 0 contributions2025-06: 0 contributions2025-07: 15 contributions2025-08: 318 contributions3182025-09: 100 contributions2025-10: 20 contributions2025-11: 0 contributions2025-12: 0 contributions2026-01: 73 contributions2026-02: 43 contributions2026-03: 63 contributions2026-04: 34 contributions2026-05: 55 contributions2026-06: 17 contributions2026-07: 18 contributions2026-08: 171 contributions2026-09: 1193 contributions1,1932026-10: 0 contributions

S1Economies

Agent-based economic simulation · since Aug 2026 · Active

Econ-Engine

A deterministic agent-based economy in the browser, built so that every outcome can be traced to a cause.

S1Subsystem · 3 projects

Economies

What happens to inequality when you change the rules?

Econ-Engine dashboard showing household, market and government metrics
Fig. 1Econ-Engine running in the browser: seeded, paused on day 0, $5,000 in circulation.
$5,000CONSERVEDpurchaseswageswealth taxtransfersprofit taxHOUSEHOLDSFIRMSGOVERNMENT
Fig. 2Econ-Engine’s circuit. Households buy from firms and are paid wages; the state taxes wealth and profit and returns every cent as transfers. It can neither borrow nor print, so total money never changes.
Agent-based economic simulationSince Aug 2026Active

Econ-Engine

A deterministic agent-based economy in the browser, built so that every outcome can be traced to a cause.

households, seeded and persistent
100
households, seeded and persistent
in circulation, conserved to the cent
$5,000
in circulation, conserved to the cent
wealth taxes tested against inequality
0–100%
wealth taxes tested against inequality
S2Subsystem · 5 projects

Strategy & worlds

Can a machine learn to play a world, and can a world be modelled well enough to play?

genuine 11×11 maps in a hash-verified split
5,517
genuine 11×11 maps in a hash-verified split
globally stable action IDs with legality masking
63,913
globally stable action IDs with legality masking
values per observation under human-information parity
6,424
values per observation under human-information parity
semantic features per legal action
47
semantic features per legal action
Fig. 3PolyVision in four numbers: real maps, a stable action vocabulary, what the agent sees, and how each legal move is described.
63,913ACTION IDSlegal actionnext state + reward6,424 valuesPPO POLICYGAME ENGINEOBSERVATION
Fig. 4PolyVision’s learning loop. The policy picks one of the currently legal actions; the Java engine plays it and returns the next state and a shaped reward.
Reinforcement-learning research environmentSince May 2026Active

PolyVision

A research environment for training and benchmarking reinforcement-learning agents on The Battle of Polytopia.

Began from ClaireBookworm/polytopia_rl and has since been substantially extended.

Top-down procedurally generated hotel floor plan with corridors and rooms
Fig. 5Un-Dead Hotel 2D: a procedurally generated floor of Hilbert’s.
Colony survival gameSince Mar 2026Prototype

Un-Dead Hotel

Colony survival inside Hilbert’s: an infinitely generated hotel overrun by zombies. A 2D browser prototype, now being rebuilt in Unity 3D.

JavaScript / Unity / C#

Geopolitical world-state baselineSince May 2026Active

Atlas Core

A province-level model of the world on 1 January 2025, assembled from public datasets as a foundation for strategy games and simulations.

Country data on health systems, governance, trade and security is merged into canonical datasets alongside GHSL settlement statistics and Natural Earth infrastructure, all inspectable on a MapLibre map.

React / TypeScript / MapLibre GL / GeoJSON

S3Subsystem · 8 projects

Artificial life

How much behaviour can emerge from simple local rules?

Coloured particles forming clusters, with an interaction matrix panel
Fig. 6Particle Life: 1,800 particles, six types, one interaction matrix.
Clusters of violet particles on a black field
Fig. 7Particle Simulator: gravitational clustering under Barnes–Hut.
sight, soundthreatsrecallgoalbuild, gatherhunger, sleep, coldWORLDPERCEPTIONMEMORYBEHAVIOUR TREEACTIONDRIVES
Fig. 8One simulated human, every tick. Drives from modelled physiology push the behaviour tree; what the agent does changes the world it perceives next.
Artificial-life simulationSince Jun 2026Active

Life Engine 3D

Autonomous humans in a 3D world who perceive, remember, get hungry, get cold, and build shelter to survive.

visual field with line-of-sight raycasts and hearing
200°
visual field with line-of-sight raycasts and hearing
physiological drives, adenosine and ghrelin, modelled over game time
2
physiological drives, adenosine and ghrelin, modelled over game time
maximum accelerated timescale with live behaviour-tree debugger
64×
maximum accelerated timescale with live behaviour-tree debugger
S4Subsystem · 5 projects

Personal systems

What if a life were run on evidence rather than impressions?

Fabbro Systems: three applications with explicit boundaries, one shared design system, and a rule that observations are never mixed with interpretations.

Kleos landing page with a radar chart scoring eight life vectors, one marked unknown
Fig. 9Kleos: eight vectors scored from evidence. Where there is none, the answer is “unknown”.
Heracles landing page with a phone mock-up of a live workout
Fig. 10Heracles: the public surface of the training ledger.
Ariadne landing page reading “Turn direction into action”
Fig. 11Ariadne: the thread between direction and today.
training logsstrengthcurrent stateprioritiesnew datahealth, recordsRREINFORCINGEVIDENCEHERACLESKLEOSARIADNEACTION
Fig. 12Fabbro Systems as one feedback loop. Heracles and other evidence feed Kleos’s model of the current state; Ariadne turns it into priorities; acting on them produces new evidence.
  • Heracles

    Open ↗

    Measures one domain deeply

    An evidence-aware strength-training app: logs every set and shows what it says about strength and which muscles it trained, with the working shown.

  • Models the whole current state

    An evidence-based model of a whole life across eight vectors, stored as auditable, dated snapshots with explicit coverage and confidence.

  • Ariadne

    Open ↗

    Turns direction into action

    A personal strategy system linking long-term direction to today’s tasks, with an AI assistant that reprioritises by strategic relevance.

Lineage

  1. 2025-08Stasis
  2. 2026-04Blackbox
  3. 2026-04Ariadne
  4. 2026-08Heracles
  5. 2026-09Kleos
S5Subsystem · 44 pieces

Institutions

Which of the systems we live inside are broken, and what are the levers?

Eight essays to start with

  1. 01To Govern A Garden
  2. 02Democracy Needs Dangerous Men
  3. 03Your Panic Is For Sale
  4. 04The Rent Control Paradox
  5. 05After The Factory
  6. 06The Unwinnable Game
  7. 07Haunting the Market
  8. 08Your Data Is Useless

Argued here, built there

30 Jul 2025 → 11 Aug 2026

The Speculator’s Game → Econ-Engine

Argued about taxing wealth and speculation in 2025; tested wealth taxes against inequality in simulation in 2026.

25 Nov 2025 → 22 Apr 2026

The Mandate for Speed Bumps → Blackbox

Argued for deliberate friction against manipulative design; built a task app around friction a few months later.

§ 02Trajectory

Compounding.

The same instinct at every stage: find a system, get inside it, and build something bigger than the last time.
  1. AGE · 2018

    14

    Twitch affiliate

    and a summer course in machine learning at Imperial College London.

  2. AGE · 2020

    16

    204,000 followers

    on TikTok, grown from 6,000 in nine months.

  3. AGE · 2021

    17

    A political community

    built on Twitch around ideas, with its own website.

  4. AGE · 2025

    21

    34 essays, 7 columns

    in eight months, and the first simulations.

  5. AGE · 2026

    22

    Five subsystems

    PolyVision, Econ-Engine, Life Engine 3D and the Fabbro Systems apps.