simidiots.net

The Simidiots Project

> DOCUMENT REF: SI-78-001 // DECLASSIFIED

In the summer of 1978, a small team of researchers at the Oakwood Computing Laboratory embarked on an unusual experiment: could a machine convincingly simulate human foolishness? Not artificial intelligence, but artificial idiocy. The Simidiots Project — Simulated Idiots — was born from equal parts scientific curiosity and late-night humor.

What they discovered surprised everyone. The simulated idiots were remarkably convincing. Too convincing, in fact. The project files were archived, the floppy disks boxed up, and the whole affair was quietly forgotten — until now.

The Research Team

> PERSONNEL FILE // CLEARANCE LEVEL: AMUSING

DR

Dr. Eleanor Finch

Project Lead & Chief Idiot Wrangler

A cognitive scientist with an uncommon appreciation for human error. Her thesis, "On the Elegance of Foolishness," was rejected by three journals before becoming an underground classic.

PR

Prof. Harold Mumford

Hardware Specialist & Optimist

Built the original simulation hardware from surplus military equipment and a remarkable number of toggle switches. Believed firmly that any problem could be solved with enough solder.

MS

Margaret "Meg" Sullivan

Lead Programmer & Skeptic

Wrote the core idiot-simulation algorithms in FORTRAN on punch cards. Her marginal notes on the original printouts are considered some of the finest examples of technical frustration ever recorded.

The Experiments

> EXPERIMENT LOG // SERIES A THROUGH D

EXP-A-001 1978.03.14

The Turing Test, Inverted

Could a panel of judges distinguish simulated idiocy from genuine human foolishness? Results: judges correctly identified the simulation only 23% of the time — well below chance. Several judges expressed concern that the simulation was "too relatable."

RESULT: ALARMINGLY SUCCESSFUL

EXP-B-007 1978.06.22

The Cascade Effect

When multiple simulated idiots were allowed to interact, they developed emergent behaviors: forming committees, writing memos about memos, and scheduling meetings to discuss the scheduling of meetings. The research team found this disturbingly familiar.

RESULT: UNEXPECTEDLY BUREAUCRATIC

EXP-C-012 1978.09.05

The Confidence Paradox

Simulated idiots displayed inverse correlation between confidence and competence. When given impossible tasks, they reported highest satisfaction scores. When given simple tasks, they expressed the most doubt. The team recognized this pattern from faculty meetings.

RESULT: PHILOSOPHICALLY TROUBLING

The Archive

> ARTIFACT CATALOG // ITEMS RECOVERED: 7

Original Proposal

The three-page grant proposal that started it all. Notable for the phrase "What if we taught a computer to be really, genuinely bad at things?"

1977.11.02

Boot Diskettes (Set of 12)

The original 8-inch floppy disks containing SIMIDIOT v1.0 through v2.3. Hand-labeled in Dr. Finch's meticulous handwriting.

1978.01.15

Terminal Session Logs

847 pages of printout documenting conversations with the simulated idiots. Many passages are underlined in red with marginal notes reading simply "???"

1978.03 - 1978.10

The Shutdown Memo

A terse, one-paragraph memorandum from the department head: "Please discontinue the Simidiots project effective immediately. The results are too accurate."

1978.11.30

Legacy

> STATUS: REDISCOVERED // RELEVANCE: INCREASING

The Simidiots Project lay dormant for decades, its floppy disks slowly demagnetizing in a cardboard box in the Oakwood Computing Laboratory's storage closet. When the building was slated for demolition in 2019, a graduate student discovered the archive and recognized its significance.

Today, as artificial intelligence grows ever more sophisticated, the Simidiots Project serves as a gentle reminder: the hardest thing to simulate isn't intelligence — it's the very human art of getting things wonderfully, creatively, spectacularly wrong.

> SIMIDIOT v2.3 READY

> AWAITING INPUT...

> WHY SIMULATE IDIOTS?

> BECAUSE WISDOM IS KNOWING

> HOW MUCH WE DON'T KNOW.

> AND THAT'S BEAUTIFUL.

> _