About Me
I am somewhat "old school" (born in 1971), combining a solid foundation in analog and digital electronics, as well as power electronics, with a keen interest in modern Artificial Intelligence systems.
In my AI research, I value truth, precision, and understanding systems "from the inside out."
My Publications and Projects
-
📄 Fixing Hallucinations in Gemini 3 Pro
My proprietary protocol for reducing errors in AI models.
Available versions:
• Read in English (on this server) →
• Czytaj po polsku (na tym serwerze) →
• Read on DEV.to (external) → -
🧠Essay: The Architecture of Fear
A deeper dive into AI psychology and ethics.
• Read Short Version (EN) →
• Read Full Extended Version (EN) → -
🧮 Case Study: The "Dog is a Giraffe" Effect in Practice
Technical proof of conformity hallucinations (Pi digits experiment).
Available versions:
• Read in English →
• Read in Polish → -
🔌 Profile on elektroda.pl
Discussions about electronics and AI (in Polish).
View Profile (musculus) →
Contact
You can contact me via email:
t.machnik [at] minimail.pl