Tomáš Mikolov is a Czech computer scientist who, in 2012, asked what seemed like a deceptively simple question: can a computer learn that words have meaning, not just sound? The answer he found, together with colleagues at Google Brain, would change natural language processing forever. He did it without a huge team, without dozens of backing papers, and with an idea that at first many people didn't quite know how to value.
In 2013 he published Word2Vec: a technique that converts words into numerical vectors so that words with similar meanings end up close to each other in mathematical space. The demonstration that left everyone speechless was this equation: "king" minus "man" plus "woman" equals "queen." A machine that had never "understood" anything suddenly seemed to reason about language. The paper was initially rejected at a major conference, but circulated on arXiv and became one of the most cited works in machine learning history, with more than 52,000 citations. In 2023, ten years later, it received the NeurIPS Test of Time Award, recognizing exactly that: ideas that are still shaping the field a decade on.
After Google Brain, Mikolov moved to Facebook AI Research (FAIR), where in 2016 he co-published fastText, an extension that also captures word morphology and supports 157 languages. He then returned to his native Czech Republic to research at the Czech Institute of Informatics, Robotics and Cybernetics (CIIRC), and in 2024 co-founded BottleCap AI in Prague, where he works today on efficient reasoning models that bet on architectural innovation over raw scale.
Mikolov's lesson is about the elegance of simple ideas. Word2Vec wasn't the biggest or most complicated model; it was a clear intuition about capturing a word's meaning through the company it keeps in text. Every time you write to a language model and it "understands" what you want, there is an echo of that idea working underneath. Sometimes the revolution doesn't arrive with thousands of GPUs; it arrives with an equation that makes everyone say: why didn't we think of that sooner?
Official links for Tomas Mikolov, The man who taught machines to understand words
More figures who shaped AI in AI Legends, or back to the news.