Make Your Own Expert

Sometimes it’s hard to see the big picture and other times, it’s difficult to remember all the details. It all depends on which way the pendulum is swinging. Thank goodness the machines can remember it for us wholesale (that’s a nod to PKD, not plagiarism). What we mean to say is, today you don’t have to choose and if you can’t pick one, well by golly the machine can probably do both for you.

What we are talking about is relying on our computer counterparts to sum up ideas, and also break a larger idea down to it’s smaller parts to better explain how they fit together.

We can currently create a multi-modal AI model, give it a variety of notes and ideas in various formats, such as text, images, videos, and code. The machine can analyze the data and resources available and provide solutions to the problem, most likely from any perspective or context you would like it to use.

Here we are going to look at the efforts of Google to build some tools to help with learning, summation and presenting a subject or dataset to the user in a way they can understand. Sometimes that function return looks like a pie graph, other times maybe it sounds like a podcast…

And that brings us to the first tool we will be looking at, NotebookLM.

NotebookLM was initially called Project Tailwind, but was announced and introduced to the public as NotebookLM on the Google Blog on July 12, 2023. Here’s what the people at Google had to say about it on their blog (credits list …

It’s our endeavor to reimagine what note taking software might look like if you designed it from scratch knowing that you would have a powerful language model at its core… – Google NotebookLM Blog

What is it?

Ok, so if you are having difficulty understanding a topic one of the best ways to learn about it is to talk about it with someone who knows more about it. The Socratic method (a conversational question and answer system) is still an extremely valid way to learn even a couple thousand years since we gave the idea a name.

That’s where NotebookLM comes in handy. Now we aren’t talking to ourselves anymore, but communicating with another entity that has knowledge and can share it. Again, here’s some clarity from their blog:

“Think of it as a virtual research assistant that can summarize facts, explain complex ideas, and brainstorm new connections — all based on the sources you select.” – Google NotebookLM Blog

Basically, you can collect, link and upload information about any given topic or event and the program will group the data together for the AI to reference, this is called a “Notebook”. After you provide the source material and create a notebook, the AI will absorb the information and then attempt to use it to summarize the information or draw conclusions from it.

How it Works

We don’t have any kind of behind the scenes access the THE ALPHABET, but if we had to guess what the insides look like when they turn information into answers, then probably something like this…

NotebookLM is able to use RAG to get the data from the notebook’s supplied content, and then use a language model to parse and understand the user request. Once the answer is known, then the output must be shaped and formatted to be of use to the end user, so the output process works in reverse and the ML and LM determine what to share and how to do so. The program will then present their results and conclusions to the user, hopefully in a format suitable for the user to consume.

Try This At Home…

You can create you own notebook and we highly encourage you to do so. It really helps to hear another perspective on whatever you are learning. The notebooks can be used by a variety of users and formatted into different media types.

To create your own notebook, visit the following link: NotebookLM

So far most of the attention that NotebookLM has attracted stems from people using the app to create podcasts with their notebooks. We will share some of our results here, so you can get a peek, without getting your hands dirty.

Q&A Time

So we are curious have you used NotebookLM yet? If so, what for? How did it work out? If you haven’t used it personally, but know someone who has, tell us their story. If you’re not into telling stories and you have a good idea on how to use it, shout it out.

Come on. We don’t have all day, we can see The Singularity from here.