True. The title is just clickbait and what we find inside is suggestions for dimensionality reduction by a person who appears to be on the verge of reinventing autoencoders disguised as neural hashes. Is it a mere coincidence that the article fails to mention autoencoders?