Menu
AboutModelsLeadershipResearchSafetyJoin

A NATURAL STUPIDITY LAB

In pursuit
of AGS.

Artificial General Stupidity.

Human-level errors. Superhuman confidence. We are building models that can be wrong about everything, at once.

Explore our researchMeet JeanClaude
FIG. 01 / THE CONFIDENCE SCALING LAW
Confidence scales. Accuracy does not.As compute increases, confidence rises toward Ready to replace your team while accuracy falls toward a line labeled Facts became a bottleneck.MODEL CAPABILITYMORE COMPUTE →Ready to replace your team.ConfidenceAccuracyFacts became a bottleneck.
Peer review: replaced by applause.
ARTIFICIAL GENERAL STUPIDITYFRONTIER MODELSRESPONSIBLY MISALIGNED

01 / OUR MISSION

Intelligence is becoming abundant.
Stupidity remains poorly understood.

Our objective is to close that gap. We study cognitive failure as a generalizable capability: from an unsupported assumption to an entire institutional strategy.

Read our manifesto

JEANCLAUDE / THE DYSTHROPIC APP

More certain.
Less correct.

Whisper 7.3. Ramble 9.1. Rant 12.8. Rumor 15.4. Four models. One remarkably consistent absence of judgment.

Explore our models
CONVICTION↑
UNDERSTANDING—

THE TEAM

Exceptional people.
General mistakes.

Research. Operations. Finance. Brand. Product. Five disciplines, united by a conviction that the evidence will eventually catch up.

Meet our leadership
Mario DemoDay, Manuela Manny DemoDay, Hare K, Blair Hypecycle and Riley Featureflag in conversation around a table at Dysthropic

OUR INVESTORS

Backed by institutional conviction.

Capital for the next frontier of cognitive failure.

a16zillion
Conviction before capability.
SoftBrain Vision Fund
A long-term vision. A short attention span.
Seekoia
Higher conviction. Elevated returns.
DumbRock
Every asset. One collective assumption.

Research at the frontier.

All research
THE AGS THESIS / 01

Machines of Unwarranted Confidence

How Artificial General Stupidity Could Transform the World for the Worse

ResearchCAPABILITIES / 02

The Scaling Laws of Confidence

What happens when certainty scales faster than understanding?

ResearchSAFETY / 03

AGS Safety Evaluation

Toward responsible deployment of irresponsible conclusions.

ResearchEVALUATIONS / 04

The Epistemic Benchmark

Measuring the distance between what is known and what is announced.

Research

FOLLOW THE LAB

Stay ahead.
Of the evidence.

Your email is stored only for Dysthropic updates. Never sold. No emails are sent automatically.