Star on GitHub

About

The Model

The simulation uses an empirically-grounded model. Output emerges from per-task speedups (not global multipliers), skill-biased effectiveness, and validation overhead.

See Mathematical Model → and Animation Concept →

Per-Task Speedups

TaskAI SpeedupSource
Boilerplate code1.55xPeng et al. RCT: 55.8% on simple tasks
Writing drafts1.40xNoy & Zhang: +40% below-median
Debugging1.30xEstimate (no RCT data)
Custom code1.00xMETR: 0% for experienced devs
Validation0.90xModeled: hallucination overhead

Skill-Bias Finding

AI helps junior researchers far more than senior ones. Low-skilled researchers get full speedup; experienced researchers see ~0% gain.

Expected Output

Instead of 3-5x, the model produces ~1.5x output factor — more honest, grounded in published studies.

Data Sources

The Animation

The animation is a stochastic Petri net: nodes represent research states, events fire probabilistically, tokens carry confidence that increases as work is validated. See Animation Concept → for full details.

Productivity Multiplier a Myth?

Vendor studies claim 10x-55x speedup from AI. But independent research tells a different story:

StudyFindingConfidence
METR 20250% speedup for experienced devsHIGH
Peng et al. 202355.8% faster on simple tasks onlyHIGH
Noy & Zhang 2023+40% for junior, ~0% for skilledHIGH