#Embedding Vector Search vs Index Search? Difference and limitations?

3 messages · Page 1 of 1 (latest)

chilly patio
#

Hello guys, can anyone enlighten me on the difference between vector search and sorting base on cosine similarity vs index search as seen in this youtube video https://www.youtube.com/watch?v=y2z7KNcysFo&lc=UgzUTRDbCCENDLHVpSd4AaABAg ?

"In this video, we explore the capabilities of GPT in retrieving books mentioned in a podcast episode. Using Lex Fridman's podcast as an example, we demonstrate how GPT can extract book recommendations from audio. Say goodbye to manual searches and hello to the future of book discovery. Watch now to see how GPT can revolutionize the way you find...

▶ Play video
real sleet
#

They are the same thing. He is using whisper to transpose video to text, but then they are both using vector search. He is just using a different tool to do the same thing

chilly patio
#

Interesting, in the video it didn't look like he had any vectors as he used index.query instead of openai.embeddings_utils