# Rafał Rybnik

> Praktyk AI w obszarach, w których nie powinien działać — badania rynku, edukacja, organizacje pozarządowe.

## Teksty

- [Kleszcz, miliarder i drabina Rube Goldberga](https://rybnik.to/kleszcz-miliarder-i-drabina-rube-goldberga.md) — 2026-05-21: Esej o rozpoznawaniu fałszywej złożoności w narracjach konspiracyjnych. Każde ogniwo z osobna prawdziwe, łańcuch — nie. Kleszcz Lone Star, Bill Gates i alfa-gal jako wygodny przykład maszyny narracyjnej, która imituje analizę systemową. [PDF]
- [How to market to smart people](https://rybnik.to/how-to-market-to-smart-people.md) — 2024-08-14: Why discerning, sceptical, well-informed audiences need a different marketing approach — and how to earn their attention.
- [Introduction to Fourier analysis of time series](https://rybnik.to/introduction-to-fourier-analysis-of-time-series.md) — 2021-01-28: Using the Fast Fourier Transform to uncover cycles and structure in time series, with visualisations and Python code.
- [How to predict solar energy production](https://rybnik.to/how-to-predict-solar-energy-production.md) — 2021-01-10: Predicting photovoltaic energy output from weather data: a practical, data-driven approach with Python. [PDF]
- [Measuring forecast accuracy](https://rybnik.to/forecast-evaluation-statistics-with-examples-in-python.md) — 2021-01-03: Choosing, building and interpreting forecast error metrics for time-series predictions — with worked examples in Python.
- [Conjoint analysis: Discrete choice modelling in market research](https://rybnik.to/choice-based-conjoint-analysis.md) — 2020-11-20: Choice-based conjoint analysis (discrete choice modelling): estimating which product attributes actually drive consumer choice in a marketplace-like setting.
- [Conjoint analysis: How to develop the perfect product](https://rybnik.to/how-to-develop-perfect-product-using-conjoint-analysis.md) — 2020-11-09: Using conjoint analysis to design better products by decomposing customer preferences into the value of individual attributes and their levels.