in 30 years, abc.xyz renames itself to a 3000-dimensional concept-vector; translated to English, will read like dada poetry
Conversation
Replying to
automated fast-food restaurants of the future will accept custom orders as high-dimensional vectors describing quantities, heating, & arrangement of their brand's core ingredients
Quote Tweet
4
1
4
…and so it (👆) continues: fox5ny.com/news/mcdonalds
"Are you ready for the Land, Air & Sea? It combines a Big Mac, a McChicken, and a Filet-O-Fish into one giant sandwich."
Replying to
…or perhaps the evolved intermachine language won't even sound like dada poetry, when reduced to human phonemes, but rather glossalia or scat:
Quote Tweet
A known limitation of DALLE-2 is that it struggles with text. For example, the prompt: "Two farmers talking about vegetables, with subtitles" gives an image that appears to have gibberish text on it.
However, the text is not as random as it initially appears... (2/n)
Show this thread
1
2
this guy gets it: research papers of the future will be giant raw dense embedding vectors. if you want to read them in slow, low-density human language, feed them into your local LLM.
Quote Tweet
Replying to @pmarca
It looks inefficient to expand using gpt and then read summary using gpt when you consume papers. Authors can just publish an embedding vector with numbers instead of words. You paste these numbers to gpt and then consume this knowledge how you like it.
2
9