The Prompt Is Not The Problem, Your Files Are

AI doesn’t “make things up” because it’s evil. It makes things up because we keep asking it to perform miracles without giving it our real context. We dig into why the most valuable lever in artificial intelligence is your knowledge base and how that single shift changes everything from accuracy to speed to brand voice.

We compare the strengths people associate with today’s big AI chatbots: ChatGPT for planning and writing, Claude for coding, Grok for real time awareness, and Gemini as Google’s fast improving assistant. Then we zoom in on the move that matters for anyone running a business or team: Google Gemini being baked into Google Workspace. When your emails, docs, meeting notes, calendars, SOPs, and assets already live in Gmail, Drive, Docs, and Slides, your “source of truth” is sitting right there.

NotebookLM is the clearest example of this approach. You can upload the exact sources you trust, ask questions against them, and get answers with citations so you can verify in seconds. We share a simple five part framework for building a powerful AI library: your core thinking, your customer language, your proven content, your process documents, and outside trusted sources you approve. If you want better AI outputs, start treating your documents like an asset.

Subscribe, share this with a teammate who owns marketing or ops, and leave a review with the biggest challenge you have building a clean knowledge base.

Transcript

Google Workspace And Captured Audience

Gemini Gets Baked Into Everything

NotebookLM And Grounded Answers

Five Pillars Of A Knowledge Base

Leapfrogging With A Business Brain

Newsletter And Help Offer

SPEAKER_00 0:06

We're live. I made simple. We've been doing a lot of conversations around Notebook LM. And as you know, as I've been watching Google change um or update, upgrade the products. You know, what we saw was uh chat GPT made a big splash and got everybody um got everybody's heads around what's the expectation expectation between a chat bot, um uh an AI chatbot. And so that was like everyone's uh or most folks introduction to understanding large language models and AI, helping you produce content, thinking, strategy, etc. In the meantime, other products such as Claude or Grok and uh Jim and I have been um you know making their move, positioning their products to really excel in what they bring to the table better than the other folks. Grok, for example, that's XAI's um chat bot. It brings in real-time, you know, X is basically a group chat, a global group chat in the moment, updates, people talking about all the perspectives about any topic you want to know. And Grok brings in their their knowledge through kind of uh having a finger on the pulse of the world, on the topics, on whatever's going on. And that's what's the strength that Grok brings. Um, Claude, Claude has its um perspective, and it seems to be like uh folks love it for coding, it's like one of the best coding tools. Chat GPT is um great at maybe uh planning or writing as well. But Gemini, I've been watching Gemini, is like you would expect that Google would have had their act together and would have been leading this charge. But um I think in the background, what they've been doing is, you know, I I asked this question to Gemini. I'm gonna share my screen here. I asked this question to Gemini. If I can find it. Yeah, so it's like I was curious about how many folks are using Google Workspace or some form of uh Google product, Google Docs, Google uh Gmail Sheets, slides, etc. And it says about it has about three billion active users. And then as far as paying business customers, 11 million using the products like um Google Workspace, which is includes Google Drive, Gmail, Google Docs, Sheets, Slides, Calendars, and Meets. And then what does that mean? It means that it means that they have a captured audience. They um they have a way to when they bring their product in, they have a way to immediately impact these folks right away. And so we see that happening. We see that happening when we go and look. Yeah, when we go and look at um their Google Workspace products, you know, they have they have all of these products right here. Gmail, drive, me, chat, calendar, tasks, sheets, slides, forums, sites, docs. They've got their Gemini app, they got Notebook LM, which we've been talking about. They've got a video editor, they've got uh Google Keep for digital notes, and then app sheet and more. And when you go and look, they have been embedding in the past six months. You'll start to notice that they're embedding Google Gemini. That's this icon here. They've been embedding Google Gemini into all of these products. And all right, we we kind of expect that they would do that, but they've really turned up the heat here the past um 60 days. And so you can bring in and get more help and support. But the point that I'm wanting to connect here is that you know, the most valuable aspect of anything that you're building inside of um an LLM is the knowledge base. And notebook LM. Really, when you start using it, and I've been using it for about four years, but the light bulb really started hitting me here the past, oh, I don't know, six months is that the biggest lever that you have in AI is your knowledge base. You hear folks always complain about, oh, I tried chat GPT one time and it made something up. And they were offended by it. But what they're doing is they're going, hey, here's I'm gonna ask you a question. I'm gonna give you as very little context. I'm not gonna give you access to any of my stuff. And I'm expecting you to read my map mine and come back with everything um correct. And what's it have to do? It has to go out and and guess what it is that you're wanting, and then try to uh combine an answer and bring it to you. And so a lot of folks, when they started jumped in the chat GPT, they were seeing that, and it what that does it it creates a um oh it creates a distrust of the LLM, and so they start to get um they'll use it reluctantly, but then they'll look at it with real critical eyes for and they should, you don't want to present stuff that's incorrect, but they'll look at it and it's like they're there so it slows down the adoption in the meantime. Google's been coming around the corner and they've been baking it into all the products. And guess what? What lies in if you use these products, Gmail or Google Drive or Google Calendar Docs Sheets, what is in this? This is your information, this is your knowledge base, this is your standard operating procedures, your meeting notes, your you know what uh who you're talking to, your Google is bringing uh Gemini in there to start pulling from from your source of truth, where you operate, where you work. And then when we look at notebook LM, I'm gonna X out of here. So when we look at Notebook LM, this is where where the concept really starts to sink in. Over on the left, if you haven't been in Notebook LM, when you set up a new a new notebook and you pop one in, you set one up, it's gonna ask you to drop in your files. And so you can bring in and upload any kind of files, YouTube videos, your website, your Google Drive documents. You can copy and paste text and bring it in. You can send it out on the web to go out and search through your drive through the web if you want. You can even send it to go do deep research. But what's happening is it's saying, I'm we're only going to operate with our data sources over here, your knowledge base, your source of truth. Whatever you bring in here is we're only going to refer to it. And so when you come in and ask the chatbot, then it's only going to look over here and it's going to pull up whatever the answers are. So when we go and we look at today's episode, this is the notebook for today's episode. We brought in my newsletter. We brought in the topics that I want to cover, which is um, which is what we're talking about today. So this is my newsletter that will go out to my audience, and it talks about the knowledge base and how it's a big lever. So this gave me a source guide. This is a source guide. This is what it says. And and so now when I ask it, when I interrogate the source, you can see that it pulls up where it got this answer from. All right, so back to here. So when Google is using your information for that knowledge base, it's doing the same thing, it's bringing in data from somewhere in your workspace. And then when you get your output, it's from your source of truth, it's from your knowledge base. And then when you think about all right, now if I want this tool to really uh up-level my output, and the beautiful thing about notebook LM is basically now you have a whole marketing uh creative team that can produce any assets you need to communicate clearly to your customers, to your internal team, to your um search engines, whatever that may be that you need to do, you now have these um story, fine-tuned, super powerful asset uh generators that are pulling from your source of truth. So where your leverage comes in then is in knowing knowing what what to put in, that becomes the biggest lever for you. All right. And so you're you're going from general AI to a now focused source of truth. You can build your own library that you would have this tool, notebook LM or Gemini, comb through and go through your information that's grounded and cited, and it becomes very strong and it becomes your unique, your unique um um perspective on the world, so that when your content is is generated from your source, then now your output is going to be extremely focused, extremely dialed in. It'll represent your voice, it'll represent your brand, it'll represent your unique perspective on the way that you see the world and how your solution impacts their future. And so you have these five pillars. I like how this said, you know, you're building your library, and so we have these five pillars of a library. Look at this. This is the slide generator and notebook LM, but you can see how it's building your visual concepts based on based on this one little knowledge, uh, one little source that I uploaded in here. This is what it um output. So when you're building your notebook LM, you come in and you you bring in these sources, these are this is what you need to think about that you add to this so that when your output um when you request a a uh video overview or an infographic or a slide deck that you need to present in a sales presentation or to an investor, then these items will be influenced by these things. All right. So, what are the five things you need to be putting in here? Your core thinking. You have frameworks, you have offers, you have positioning, you have strategic ideas already in some form, in some document, in some notebook somewhere in your files. So when you bring those in and upload them as a source, now it's going to start impacting and representing your core thinking, your customer language. If you have sales calls, you have testimonials, you have reviews, um, you have all sorts of uh input coming from customers. These are where you can bring in your customer language. You know, oftentimes we're too close to our product, we speak in lingo, but our customers speak about our products in just normal language. If you go and collect all of that information and bring it in, then you're creating your customer language. You're you're getting, you're having an expectation now that um the output is going to represent you in your customer's language because you had you brought it in from all of these sources. Your proven content, you probably have newsletters or high-performing blog posts or presentations or training materials that you've invested in. This would be another source that you want to bring in as your proven content. You have onboarding documents, you have internal playbooks, standard operating procedures, checklists. These are your process documents, and you can bring those in. And now you've sit you're telling um your um project here, your notebook LM. This is how we operate. This is examples of of high performing content. This is how we think, this is our products, this is our our name, pricing, everything. This is how our customers speak about us. And then you can go out and bring in any outside trusted sources. So maybe there you have you have um organizations that you belong to, or you have um a book, or you have videos, or whatever exists out here, industry reports, research, white papers, strategic articles, you can bring those in. And now you've brought in you brought in information from outside that refers to you in the way that you approve. And by bringing in these five sources, these versions of sources into your notebook LM, then you've created a powerful source of truth that when you come in and you use that chatbot, then it's not going to bring you BS. It's going to bring you, it's going to be inculcated. You know, when you hire someone and they it takes them years to really, if you think about it, it would take them years for someone to come in and go, all right, to understand your framework, your products, your customers, your pricing, your customer language, how people refer to you. You know, how long would it take someone to come up and be um clear and familiar and confident and knowing where all this information uh comes from? It takes years, right? And so by bringing these items in, you just leapfrogged and you brought in a brain that now you and your team can refer to, and it pulls information that is in your brand voice that understands your mission, your vision, and it can produce content that you can be uh confident with. And so, you know, when you're thinking of, all right, first of all, notebook LM is free, and then the pro version is only$20 a month. But if you were to come in and set up, sign up for Notebook LM, and you just brought in one document that explains how you think. Let's say you're a nonprofit, one document explains uh what your your nonprofit stands for, who you serve, one that captures your the way your testimonials from the folks that benefit from your nonprofit. Here's an article that um that talks about maybe a case study, or maybe it was in a um a shareholder or board member packet, and then your process document. Maybe this is a process that you you take your uh folks that benefit from your nonprofit. This is what it looks like, all the steps, and then an outside article that someone's talked about, a testimonial, something that um that again is an outside verification of how the world perceives your nonprofit, then you've created a brain that takes advantage of the most valuable piece of strategic leverage, and that is your knowledge base. Right now, if you were to just go to Chat GPT and ask about your nonprofit, what would it bring back? You who knows what it would brings back. You you would not be in a position of confidence necessarily, you'd really have to study it and see what is correct and what is uh wrong. But imagine being able to upload these items, and now you have a source of truth, the biggest lever available to you that helps you um excel by using these um these tools, and so notebook lm is is that place, and what you know it's using notebook lm, it really when I started seeing this trend where Google Gemini is being baked into every product, this communicates clearly that their main advantage that they have is that they understand that the most valuable piece in this whole puzzle for folks to get the best value from artificial intelligence is this knowledge base and being able to cite where that information came from within your knowledge base, and who creates the knowledge base? You do, and you can you can drive imagine if you you're investing in this resource here is going to clearly um clearly impact the output of the of the assets that it's helping you create. And so clever prompting is you know, it's nice to know how to prompt, but more importantly, it's that knowledge base. So upload your core frameworks, position to teach the AI how you think, include the customer language, your operations, your standard operating procedures, bring in proven content and research. And then your knowledge base now is your real lever. The citations eliminate any concern or any uh insecurity about is this information is correct. You're able to click on that little document that Notebook LM refers to. Even Gemini will pull from these when you go look at uh how to create a slide or a document, it will cite the information that it's pulling from. Show you exactly where it pulled the citation from. And so this becomes your source of truth. And knowing this now is like one would go through and start to clean up their documents. One would invest a little bit in going, you know, my documents are a little fuzzy. I need to, I need to get them dialed in. I need to get them more uh better quality. I need to focus more on our are we clear with our mission, our vision, our our brand guidelines, our you know, how do we refer to our products? How do our customers refer? These things begin to help you focus as a leader in your organization on where you can start cleaning up your knowledge base that's going to give you the leverage for you to excel. All right. So that's what we learned today. You know, Notebook LM and Google Gemini are starting to show us how important your knowledge base is. Stopping the share. How do I here we go? Stop in the share. There I am. All right, team. You guys uh if you need some help, let me know. Be sure to follow us and get see that QR code in that QR code. Cool. Scan that QR code, sign up for my newsletter, like and subscribe. If you need some help, let me know. We'll help you strategize about getting a knowledge base and where to put it and how to empower your team. All right. Vince, thanks for your comment. No, well, everybody's struggling with marketing tools. I agree with I agree with you, Vince. Let us know. All right. We'll see you next week on AI Made Simple. That's a wrap.