---
title: 'AI Explained: What is RAG - Retrieval Augmented Generation?'
source: 'https://youtube.com/watch?v=tKPSmn-urB4'
video_id: 'tKPSmn-urB4'
date: 2026-07-28
duration_sec: 176
---

# AI Explained: What is RAG - Retrieval Augmented Generation?

> Source: [AI Explained: What is RAG - Retrieval Augmented Generation?](https://youtube.com/watch?v=tKPSmn-urB4)

## Summary

The video explains why AI companies like Google and OpenAI are paying for data from Reddit, WordPress, and Financial Times. It distinguishes between using data for training AI models and using it as a grounded source for retrieval-augmented generation (RAG). The core concept is that RAG improves AI responses by retrieving real data from a database before generating an answer, making outputs more accurate and reliable.

### Key Points

- **Data Deals with AI Companies** [00:00] — Reddit, Automattic (WordPress/Tumblr), and Financial Times are selling data to AI companies like Google and OpenAI, which seems counterintuitive because AI often undermines these sources.
- **Two Uses for Data** [00:29] — AI companies can use data for training new models or as a grounded source for retrieval-augmented generation (RAG).
- **How AI Systems Work** [00:49] — A prompt goes into the AI system, which generates a completion. However, the AI may produce plausible-sounding but incorrect answers because it only predicts tokens, not truth.
- **Grounded Sources** [01:23] — Adding a grounded source solves the accuracy problem: the AI sends the question to a database, retrieves matching information, and combines it with the prompt to produce a grounded response.
- **Retrieval-Augmented Generation (RAG)** [02:10] — RAG retrieves information from a database and augments the AI's response with it, leading to more accurate answers.
- **Semantic Cache** [02:19] — A semantic cache stores completions so that future similar prompts can bypass the AI entirely, improving efficiency.

### Conclusion

The future of AI lies in grounding models in real data using RAG, rather than relying solely on training data. This approach makes AI responses more reliable and efficient.

## Transcript

you've probably heard the story of how
Reddit is selling their data to Google
for AI scraping how automatic is selling
all the WordPress and Tumblr data and
now how Financial Times is selling their
data to open Ai and this doesn't sound
right right because we've been told that
these AI companies are taking all this
data and basically invalidating the data
sources they're coming from by sharing
the data so you don't get to the source
so what exactly is happening here this
is more complicated than it sounds you
see there are two different things these
AI companies can do with the data one is
use it for training you know building
new models based on the data and they're
definitely doing that in some respect
but the other one is to use the data as
a grounded source and that's really
interesting so let me explain here's a
very simplified drawing of what happens
when you use a AI system like chat GPT
you put in a prompt The Prompt goes to
the AI system and the AI system creates
a completion this is the response that
comes out of the AI system the challenge
with this model is if you ask the AI
system something that exists within its
training data it's likely to put
together something that looks like an
answer that's correct but it's just
putting together tokens to make up
something that looks like language it's
not actually answering your question so
there's no guarantee that the answer
will be correct the way we solve this is
by adding a grounded source so when you
ask the AI system a question the AI
system sends off the question to a
database with information and then
matching information gets sent back into
the AI system it combines that with your
request and you get a more grounded
response that's actually grounded in
truth and this is what they're going to
be doing with these different media
organizations instead of the AI system
just straight up trying to answer the
question it'll go to a grounded source
to get some information first and use
that to process an answer in the same
way that when you're working with chat
gbt if you ask it to write an article
it'll write a okay article but if you
write the start starting point of an
article and ask it to help you make the
language better it'll do a much better
job this process is called retrieval
augmented generation because it
retrieves information and then it
augments that information that gives it
back to you and doing this we can take
things one step further and introduce
what's called a semantic cache so that
when you put in a prompt we can go over
to the grounded Source get some
information put that back Ai and then
when the completion comes out we'll set
it into the cache so that next time we
can bypass the AI entirely this is the
direction AI is currently going and will
be going in the foreseeable future
instead of using the AI system to try to
answer the question directly have the AI
system ground itself in real data and
use rag retrieval augmented generation
to pull real data out and then augment
it in its response hope that helps