After learning embeddings, everything becomes a high-dimensional vector

TheTrustyPwo · reddit · 2026-07-23

A playful post about “embeddings brainrot”: the author jokes that after learning token embeddings, they now see almost everything as an n-dimensional latent space.

They extend the joke to personality, music taste, emotions, generational trauma, and even crushes, imagining absurdly scientific ways to analyze human behavior with vectors, cosine similarity, and PCA.

Original post →

More from Fun

Fun channel →