AI criticizes my son’s first art piece
I asked top AI models to analyze my son’s first art piece. What they said says more about AI than the drawing itself.
Do AI detection actually work?
I’ve spent a year mastering AI, but I still write my own content. So why are AI detectors claiming my human thoughts are machine-generated? I put my own writing—and a few popular 'detection' tools—to the test. The results were not just surprising; they were a wake-up call for every digital creator.
Stars, Signs, and Silicon: My AI Astrology Experiment
Astrology was arguably the world’s first pattern-recognition engine. I used five of the world's most powerful AI models to analyse my natal chart and apply it to my career strategy. The results were a mix of "CEO-level" insights and fascinating cosmic hallucinations.
AI only understands English?
I put five top AI models to the test with a simple question in Serbian: "Is egg yolk white?" The results were a surprising failure for nearly every model. This personal story reveals a crucial blind spot for LLMs: their struggle with linguistic nuance and data from smaller language groups.
Math puzzle challenge AI models
I tested 6 top AI models (ChatGPT, Gemini, Copilot, etc.) with a classic "three 3s" math puzzle. The results highlight a major problem: AI models lack elegance and overcomplicate simple solutions.