CH NEO-ZÜRICH EDITION
WEATHER · CLEAR 14°C
BLEND OF THE DAY · 07/ROGUE
EST. 2027
THE AEC CYBER MORNING NEWS

PAZ Kaffi

DESIGN · DEMOLITION · CAFFEINE · DISPATCH
EDITION 0823 · 23 August 2026
BROADCAST 04:42 CET
2,400 BROADSHEETS PRINTED
READ TIME · 47 MIN
Cyberpunk Got the Topology Right — Now Draw Your Own Dependency Graph
ART
FRAME · 06:55
23-08-2026

Cyberpunk Got the Topology Right — Now Draw Your Own Dependency Graph

Cyberpunk's real prediction wasn't neon — it was topology. Read Gibson as a dependency graph and find your stack's single point of failure before it finds you.

In 1994 a doctoral student named Stephen Paul Marsh, working at the University of Stirling, did something an engineer would call reckless: he tried to give trust a number. His thesis, Formalising Trust as a Computational Concept, argued that trust was not a warm human feeling but a variable — something you could compute, decay, and revoke. It is the quiet root of what NIST codified twenty-six years later as SP 800-207: never trust, always verify, because there is no inside anymore.

Ten years before Marsh, William Gibson had already drawn the map. Neuromancer (1984) opens on “the perfect untroubled blue of a television screen, tuned to a dead channel,” and beneath that sky a world of hollow states and a handful of corporations that own the compute. As El País noted in its 40-year retrospective on the genre, Gibson had coined “cyberspace” two years earlier, in the 1982 short story Burning Chrome — more than a decade before the web made the word literal. The movement hardened fast: Bruce Sterling’s 1986 anthology Mirrorshades gave it a manifesto, and the same topology runs through Neal Stephenson’s Snow Crash (1992) and Richard K. Morgan’s Altered Carbon (2002).

Read as a network diagram instead of a mood board, cyberpunk’s real prediction was never the neon or the rain. It was the topology. “High tech, low life” is what a dependency graph looks like when every edge routes through three corporate hubs and the redundancy has been optimised away. Fredric Jameson called the genre “the supreme literary expression of late capitalism.” A systems architect would call it a single-point-of-failure diagram with better lighting.

This is not a literary aside for an architecture desk. Your own stack has the same shape. The model you author today lives inside one vendor’s cloud, licensed monthly, syncing to one proprietary format — techno-feudalism with a ribbon interface. The counter-move is old, boring, and it holds: openBIM. IFC 1.0 shipped in June 1996 as a vendor-neutral data model; IFC4.3 Add2 (2023) stretched it out to rail, roads, and bridges. buildingSMART’s argument is Marsh’s argument in another domain — the data of a building belongs to the building, not to the hub that happened to draw it.

←TODAY: Gibson’s 1984 corporate-hub world is your monthly licence and your single-format lock-in.
→3012: The offices that last keep their models in formats that outlive the vendors who wrote them.
Fulcrum: Cyberpunk read forward and zero-trust read backward name the same law — no inside is safe you cannot re-verify from outside.

Atelier: The office living with AI tooling right now is quietly re-centralising — every prompt, every render, every model routing through one API it does not control. Draw the graph before signing the renewal. The Monday move: export one live project to IFC and open it in a second, unrelated viewer (Bonsai will do). If it opens clean, you own a real exit. If it does not, you have just found your first single point of failure.

Hack: Find the single node whose removal splits your stack — the articulation point — instead of trusting the tidy architecture diagram. List your real dependencies as edges (tool → format, format → cloud, cloud → vendor) in a plain text file, one pair per line, then let graph theory surface the choke points the eye misses. The nodes it returns are the ones that take the whole office down when they blink.

import networkx as nx
g = nx.read_edgelist("deps.txt", create_using=nx.Graph)
spofs = sorted(nx.articulation_points(g))
print(spofs)

The lesson from the far side of this is plain: nobody runs out of compute. They run out of intact bandwidth and of people who remember how the old system worked, and they meet the third single point of failure on the day it fails. Michael Pondsmith, who built the Cyberpunk tabletop game in 1988, told El País in 2020, “We have become a dystopian society.” He meant the low life, not the high tech. The fix is not nostalgia — Gibson’s genre teaches that going back is already too late — it is the dependency graph you draw before the outage draws it for you. Run the four lines. Name the choke point. Then give it a backup.

Source: english.elpais.com

FILED FROM
CO-SIGNERS
PAZ Academy
CONFIDENCE
HIGH
REPRINTS
© PAZ - PARAMETRIC ACADEMY ZURICH · ALL RIGHTS RESERVED

SOURCE ·

PAZ Kaffi · multidisciplinary editorial, led by PAZ Academy

⚑ REPORT AN ERROR · SUBMIT A CORRECTION
◂ BACK TO FRONT PAGE · PAZ KAFFI

© 2026 PAZ Academy.