It takes 10 lines of code to talk to a PDF using AI with @LangChainAI

Here's the breakdown: What's going on:

- We load the PDF file
- Split the long document into smaller chunks
- Embed each chunk using OpenAI's ada-002
- Store embeddings into an in-memory Faiss Vector DB
- Ask it questions

All using Langchain's utility methods,

@pinecone just sold out their OP Stack (Open AI + Pinecone) meetup.

It took them 4 hours to reach 4x the venue capacity.

If you're even remotely interested in AI & GPT, you need to know about Vector Embeddings.

Here's why hundreds of people are rushing to get in 👇

1/6 🧵 A computer doesn't understand words the same way a human does.

LLMs like GPT-4 have their own way of making sense of words (which are technically called 'tokens').

When you say "This tastes like bubble gum!"

GPT will convert it to a list of numbers: [15, 721, 92, ...]

2/6 🧵