I gave my journals to AI, but I didn't expect this...

So I've been journaling for 15 years, and even though I do flip through my old notebooks from time to time, I would say that what I've learned from journaling up to this point has been more of a subconscious background implementation of those principles than a proactive, intentional application of lessons learned. Now, I have been getting more intentional the last couple of years at reviewing journal entries and being more intentional about how I apply those lessons in my life and work. But let's be real, there are hundreds, if not thousands, of pages to analyze, which brings me to AI. Now, there's a lot that could be said about AI, and this video is not the place to cover all of it, but one thing that AI is undeniably great at is trends and analysis. And that's exactly what I need help with right now. The main question that I'm asking myself is, what have I been writing about over the last 15 years? And how has it impacted my life and work? Over all of the hundreds, if not thousands of entries, are things actually changing for me? Or do I need to like really think more intentionally about what I'm telling myself I need to do, what I'm learning and what I'm actually doing in my day-to-day life. This video covers 4 phases of AI journal analysis and a simple way that you can start putting these practices and intentions into work in your life right away. The first thing I'm going to show you is very simply like capturing the AI text and importing it to your AI tool of choice. Now I'm not going to lie to you, this part is pretty time-consuming because the way that I did it was that I literally took pictures of every single page of not every journal, but a dozen journals. And so we're talking hundreds of pages of journal entries to analyze. Now, what I recommend doing here is just put on a movie or a show, or, you know, it's football season here in America, and so I'm watching more football. All I did was I just watched a movie. I had like my camera set up, just my phone camera set up over a coffee table, and I was just like, take a picture, take a picture, take a picture, take a picture, keep an eye on the movie, take a picture, take a picture, take a picture, take a picture. Now, AI tools are getting really good at OCR, which stands for optical character recognition. I'm pretty sure that's what it stands for. Basically like reading text that's on a picture on a screen. And so all I'm doing now is just putting these images into a specific project folder in Claude. And I decided to go with Claude because it just was processing more images more effectively than ChatGPT was. I have pro accounts on both of these tools. I find that Claude is a little bit better for this, but if you're already like dial into ChatGPT as your LLM of choice, then I think that is still a really good, a really good thing. What I would still do is create, again, create a specific project folder for this. That way you kind of keep the context to your journal and not everything else that's being talked about, like just in your regular random everyday chats. Now the next thing that you need to do is give this project folder personal context, and write an effective prompt. Now it's going to be easiest to show you this on the screen, so we're going to switch this up for just a moment, and I'm— you're going to see what I'm doing right on screen. If you want the prompt for this, all you have to do is look in the description below. It's just listed there. You don't have to sign up for anything. If you do want to sign up for something, my newsletter is listed down there as well. But if you just want to take a look at the personal context and prompt that I am using for this so that you can have an idea of what to give your own, uh, project. Go ahead and check that out. It's just linked in a Google Doc or something similar. Okay, let me show you a couple of things that I have set up in Claude, and you can do the same thing in ChatGPT also, because basically both of these are projects. And the thing that you want to give a project are like a description, instructions, and files to to like give context. And you can see here that in some ways I have a lot in here and I'll show you this in a second, but it's also not a lot, a lot. And so if you look here, this is a project for viewing and reviewing, review, no, this is a project for reviewing and analyzing journal entries. Your job is, this project is to analyze photos of my handwritten journal entries and give me helpful summaries and key takeaways that I can learn from. The other thing that you're gonna see here, and this is the, like, it's a good bit of information and context, but it's not a crazy amount. You can see here that I have an about me, like, entry, and the longer version is here in the file. But here in the instructions, just giving even more context about who I am, who I work with, who I've been inspired by, what are the values and principles that I live by, what are things that I wanna do. Every day, every week, every month? What are the things that matter to me? So that was all really organized, but I've done this over like several weeks at this point, just thinking about how I want to do this. The easiest way for you to get started with this is just to connect, have a conversation with yourself, like write a— write it out on like, I just use a text file, but basically I just talk to myself and I transcribe voice memos. And there's a great tool that I use for this called Monologue by my friends at Evry. You get it for free if you're an Evry subscriber, but Monologue is great. And then another one that I've used a lot is Whisperflow. They're really great transcription tools. And what I would say is just like, talk through your values, your principles, who you are, what you like to do, what's important to you. And then just upload that either into the instructions or you, you It doesn't have to be a Markdown file or even a text file. It can be a Word document, it can be a PDF, anything. But basically get that in here as context for yourself. So there's a lot of talk about LLMs and AI that if you give it a good prompt, you'll get a good output. If you give it a great prompt, I think that's true, but I also believe that people often overthink and overengineer their prompts. This is a shortcut. Again, shout out to Tim Grahl for teaching me this. It's like if you just have a conversation— my friend Catherine Lavery was also telling me this? If you just have a conversation as a prompt, but then keep asking this question, what do you need to know from me in order to complete this task effectively? So you can see it gives me all these questions and then I give an answer. And then I also say at the end, is there anything else you need to know from me in order to complete this task successfully? And usually you'll have to go through this 2 or 3 times, maybe more. I found usually 3 is the sweet spot. Okay, now that I've gone through this, I have a couple of quick tips to come in and tell you before we get too deep into this. The first one is that you need to have a file size that is around 1 megabyte per image. If it's much bigger than that, a lot will come in in the 3 to 5 megabyte range, even off your iPhone. And there's actually a tool for this in Mac. If you go into your folder where the images are, and go down to Quick Actions and click on Convert Image. You'll want to turn it from a HEIC. A lot of times iPhones are going to import in their HEIC format. You want to format it to JPEG, and instead of Actual Size, it's actually going to be Large, and then you'll just click Convert to JPEG. Because you're going to be asking it so much, so many questions, and because you are going to be feeding it so many images, you might hit your usage limits in the LLM a little bit faster. So again, if you have smaller files, then you won't get as close to your usage limit. But the other thing that I would do is don't use necessarily the top model for this. So if you're using Claude, just use Sonnet instead of Opus. And if you're in ChatGPT, use 4.0 instead of 5. Now the good news is both of those are still, for the most part, the default model for your LLM. And so if you don't know what I'm talking about when I say use Sonnet instead of Opus, then it's going to select that for you. Now the third phase of this is the big one, the analysis and trends. This is what we've been building towards, and I'm recording this part of the video before I actually did it. And so And what you're going to see in the next couple of minutes is my real-time reaction to what these trends analysis are. And also, and this is important, how I'm continually giving feedback to the project, to the model, trying to stay out of my own way of like crafting it to be what I think it should be or want it to be, and just let the analysis come back to me. Because again, if you've, if you've given it enough context and you have developed a, like, effective prompt that we talked about in phase 2, then this should be, you know, pretty direct and true for you. So just kind of sit back, talk to the model, talk to the project, and you're going to learn some things. You're going to see in just a moment how I'm learning some things in real time. This is my real-time reaction. So I'm reading this now pretty much in real time. So in 2011, without getting too deep into that like version of Matt. I was doing a lot of spiritual work. I grew up in the church, I'm still a part of the church. And I feel very strong in my faith. But this was a really key time when I feel like my faith was becoming my own. And, you know, we're not going to talk about that a lot. But this was a theme that ran through the journal, and spiritual source and struggle, poverty paradox. I went to Haiti in 2011, they were still really dealing with the terrible earthquakes they have, they can continually to have. Uh, yeah, it's just, it's rough there even still. Uh, this is really, this is really, uh, I was also working a lot in the Enneagram at that time. So pride, identity, true self, um, the discipline paradox. This is really interesting because this, like a lot of the thinking that I had around, uh, structure and even like time blocking. Came, came out in this time. That discipline creates freedom, not constraint. Jocko, get after it. Discipline equals freedom. So the stop doing list emerged from here as freedom from things that I didn't actually care that much about. This is actually something, and let's, let's just skip ahead a little bit. I'm just reading this for the first time. This is something that really resonates with me. Not seeking success, but significance. Not building a platform, but a life. 2011 Matt was already becoming 2024, 2025 Matt. Someone who knows that going deep beats going wide, that less is more, and the best marketing is teaching that everything flows from the well that never runs dry. The freedom you're seeking isn't found in better organization of your commitments. It's found in having fewer commitments to organize. Okay, let's talk about this. Honestly, when I saw what AI gave me for my journal entries, it was, I mean, I don't know, a little bit more emotional than I was expecting because I got this story of what things were like in 2011, the things that I was wrestling with, the things that I was going through, what my goals and dreams were. And then fast forward that to 2015, 2016, doing a vision for like, hey, what do I want things to look like in 5, 6, 7 years? And seeing a lot of those things come true. I don't think that I give myself enough— well, to say that I don't give myself enough credit, or it's like I don't celebrate or recognize the wins or the progress as much. And this is, this is a fault of ambitious people who want to do a lot of things because you're always just moving the goalposts for yourself. Looking at the analysis and trends and summaries of these journal entries, really made me appreciate how much progress I've made. One other thing that you have to keep in mind when you are talking to an AI program, the AI program is by default pretty encouraging and affirming of whatever you want to give it. You really do have to push back pretty hard and say, you need to challenge me, you need to coach me, You need to give me some things that, things that you said in the past just don't match what your actions are right now. Then you'll get much better answers and you will be able to see like where some of the discrepancies are, some of the mismatches between what a younger you was saying was important and what now the older you is maybe not paying as much attention to. In 2011, you're enough already, so live like it. In 2015, 2016, Grateful for $100K problems. I didn't earn $100K for the first time until 2019. In 2025, vivid vision of growth, a team, and $30,000 a month. Much different. The thread is still fragile. I mean, y'all are reading this, but ambition is fine. Are you still chasing enough? The risk is running past the very piece you've been chasing for 15 years. Write a definition of enough and measure it against it quarterly. If you're already living, living it, stop moving the bar. See, this is all very good. Now let's talk about this and what it means for you and what your next step is, because I went through a lot here. But do you have to do all of this? Let's talk about what I recommend. Okay, here's some takeaway tips for you as I've spent now hours uploading, formatting, analyzing, prompting, reprompting, challenging, reviewing, reviewing, and asking lots of questions of myself via AI. And the biggest First one that I would say is start with just one notebook. Like, go through this whole process with just a single notebook to start. That way you're only really talking about, say, maybe 50 images that you need to work with and one single slice of time that you will use to analyze and then like draw comparisons, see discrepancies between what you were talking about then and the way that you're living now. The second one is to make each notebook its own individual chat or conversation inside of the AI. You saw how I put all of my images and had specific prompts and everything just for the 2011 notebook, and then I opened it up a new conversation and did the same thing for my 2015-2016 notebooks. Make sure you do that. That way you can not overload a single conversation, because again, that can affect your usage limits. And it also keeps everything contextual for you. Now, the other thing that I highly recommend, especially if you are on— I don't think you can do this on free, but to me, the $20 a month is well worth it for these pro models— is to create a project for these notebooks. That way you can keep the context of all of the notebook conversations that you have in one place, and they're not affected at all or not affected very much by the other conversations that you might be having. So the context for everything stays much cleaner. So as you continue to add journal entries and add notebooks to this, one of the things that I love is that now you can use this whole project and new conversations within the project as a guide. So ask it questions about decisions that you have or problems that you're facing or things that you're interested in, because now you have hundreds of journal entries and years of personal expression that you can draw from, and that will allow you to kind of talk to yourself in a new way, in a way that honestly just wasn't, wasn't possible before. So I would definitely do that. Don't just like do the analysis and then leave this be. Use the project, use the chat, the chat feature inside of the projects to now talk to yourself and continue to use your past knowledge as a way to inform decisions and questions that you have in the present day. And honestly, just have fun with it. I think this is a really cool way to get to know yourself again and to have conversations with past versions of yourself that honestly, before this were really, really difficult to have. And if you'll allow me to like project into the future a little bit, I think this is a cool thing that eventually, you know, your kids could have access to if you're brave enough for that. And it's a way that they could, like, I think about what if your kids could have like some type of resource of like your thoughts that they could use to get advice from you directly and one day indirectly. It's a really cool thing to think about. Now, since you like this video and you're here with me, I definitely recommend this one right here. It's been super popular over the past several months. 5 ways to journal. Check it out, that way you can start to build your journal habit. There's also a free journal prompt series you can get over email. Both of those things are available to you either right here or in the description below. If you do sign up to the email prompts, reply and say hi, and I promise I'll write back to you. Okay, thanks for watching.