Transcription
If you use AI and you don't have a note-taking system, you have to start one now. And the reason lies in the clear trend that we're moving away from prompt engineering and the increasing importance of context engineering. And having used these tools for two plus years, I can really say that having the right context, having the right notes to be able to set up what you actually want, and what you want these LLMs, whether it's ChatGPT, Claude, Gemini, whoever, to be able to draw upon, the quality, quality of the output is going to be different, as well as the quality of thinking you can put into it.
Everyone's complaining that, you know, we have to outsource thinking. It doesn't have to be that way. But if you don't have notes, then yeah, there's no choice. I've talked about this in another video that you can explore here, which is it is so important to turn tacit knowledge, knowledge that just exists in our head and we cannot articulate, into explicit knowledge so that we can share with other people, which is the source of why note-taking came into place, right? Before level one, it was just us in our brain, we have things swirling in it, and because we kind of know what they are, we say, "Okay, that's fine, right? Don't need to write anything." Now, until the second iteration, we developed writing systems, and we realized, "Okay, there are things that we need to unload in order for our brains to do more." So that's why we started to translate this tacit knowledge into explicit knowledge because then we can work with other people, right? So as we're talking about on this channel, it really is very important to recognize that communication is not about performance. Of course, for for some people who are, I don't know, motivational speakers or whatever, I guess the performance side is also important. But for people who are working on ideas, knowledge workers, right? Why we need to communicate clearly is not to look good, but to get the tacit knowledge into explicit knowledge and to share with other people so they can understand, and we can take action together with as little friction as possible, right? And vice versa, when they're saying something, we can process it and understand it, which is why frameworks are really important. But that is the core of communication. And now we're adding on another layer of it's not just us and other people, but there's also an AI that has access to other kinds of information, so we can put it together, all move forward. And having a note-taking system that has the important information that the LLMs can use makes everything better. It then you can use it to work on the things that take you a lot of time. And I'll show you an example in just a moment. But just want to have a note-taking system and have information downloaded from your brain onto something that LLM can read and then use is a superpower to be able to summarize, to be able to spot patterns, all of those good stuff.
Okay, so that's part number one. Now, what do we actually put into the notes then? What kind of notes do we need to have? Well, luckily, there are just six categories we have to think about. And I'm drawing straight from context engineering because, uh, to be able to communicate with AI, we might as well use its rules and its way of thinking. So we mentioned three already that mostly are covered in prompt engineering, which is we have to give it a role, and then give it some general idea about the goal and the audience, uh, that is going to use that output. Then also, we want to give it some constraints around style or format and these kinds of things. Most people know this part already. Now, the next three things are more covered in context engineering and also notes that we want to have, which are about the inputs, the source of truth, as well as judgment.
Okay. So by input, I mean, for the LLM, you know, what's the data, just in general, that it can use. So, for example, for me, everything in my vault, in my Obsidian vault, it can use. Or maybe it's okay, maybe there are some personal things or sensitive things or confidential things I don't want it to use. Then I can also say, "Okay, use everything in my vault except for these tags, right?" You want to be able to tell it explicitly what's the input it can use. Number two is the source of truth. So, what kind of documents outrank generic information, right? So, for example, when you have a "do not use," it means yes, you have these things in your general information, but I, I don't want you to use them. That's a source of truth. So that it knows, "Okay, I'm not going to rely on those. I'm going to focus specifically on these, um, these documents." And the third one is judgment. So, how does LLM judge what is good, what is bad, right? This is especially where we need to take tacit knowledge into explicit knowledge. And luckily, there are frameworks to do this, and frameworks help us better articulate what's good and what's bad. So that's the type of notes that we're going to take.
A quick note on frameworks. I think if you've used AI for a while, then you probably noticed it has several frameworks that it keeps on using, right? For example, for formatting, it uses the heading options. So that as it goes through, it organizes things by headings. Or it likes to say a lot, "Like it is not this, it is that." Right? A great little via negativa there. But the problem with its frameworks is it keeps on using the same ones, and it keeps on using very generic ones when it produces output. But to be able for you to judge what's good or what's bad, a lot of the time, a simple way of doing that is to say, "Okay, write me a sales letter, right?" And mostly the LLM might use AIDA, like something that's very basic in copywriting. But if you have a different framework you would like it to focus on, then just share that one, right? And this way, you are inherently building what's good into, uh, your prompt and your context so that you don't have to worry about, "How does it judge?" "I don't know, like intuitively, I look at it as pretty good, right?" Instead, what you can do is give it a framework. Or if really you don't have any frameworks, but you have examples of what's good and what's bad, it would be great to have, you know, the good and the bad so that you can get your LLM to figure out what is the, what is the framework underneath. And the good thing about LLMs is they're great at distilling, right? As long as you give it information, it's pretty good at distilling. So, what you can do is if you have 10 examples of things that you think are great and 10 examples of things that you think are terrible, get the LLM to create that framework for you, right? "Help me articulate what makes these examples great and what are the patterns of the bad ones, right?" Then you've got your own framework that you can use next time consciously. So that is a great way of injecting framework in a way that helps you get the outcome that you want.
All right, I'll give you an example from my actual Obsidian vault so that you can see how not only the quality of the output changes, uh, also the quality of the thinking changes. And I really believe that yes, working with AI can absolutely make us dumb. And studies have shown, right? If we just outsource our thinking, then yes, as we go on, quality of the output drops, also we are disengaged from the thinking experience. But that's not the only way to use LLM. And studies have also found, depending on how you use it, right? You can deeply engage and actually think deeper and access things that you didn't have the mental capacity to do before. And really, what is holding us back from being more engaged in the thinking process is, have we downloaded the context out from our brain and into the LLM? So when we have a conversation, it's an intelligent one. It's one that's based on the information that we wanted to work with, right? So if you feel like, "Okay, I'm just getting dumber by the day using ChatGPT or whoever," then you have to also ask yourself, "Am I also just consuming and not processing it, outputting documents that I can feed my LLM?" Because if you're not doing that process, right, then there is not much context to share, right? Because it's difficult. You, not every single time you're there, you want to start typing out the context. It's very time-consuming. There's a lot of friction, which is why taking notes as a habit that helps you just drag and drop the context into the LLM and your conversation with it, right? So if you feel like, "Okay, this is what I'm missing," then let me just show you how significantly things can change.
For the example, I'm going to use a problem a lot of us on this channel face, which is I'm multi-passionate. I'm interested in a bunch of stuff, but I feel so scattered all the time. All I know is I don't want that singular track career or, you know, life goals that other people have. But at the same time, I don't know how do I focus in order to get to where I want to go. Like, there must be something that can tie everything together. So let's talk about that and let me show you the difference between talking to, I'm going to use ChatGPT here in incognito versus using your own context. So I'm going to share my Obsidian vault in a moment, but let's give this a try. Okay. So what we can do with ChatGPT is, I'm interested in creativity and frameworks. So the choosing contradictory, help me see how to combine these two interests together into a career that makes sense. Let's see what it says. I haven't really shared anything, but let's see what it says. Uh, blah blah blah blah blah blah. Where is the design thinking, product strategy? Okay. Creative direction with systems, creative director, brand strategist. I mean, I guess, but you know, for example, the career I'm thinking about is not like these roles, but of course, I didn't tell it. So, okay, makes sense. It's saying something like this. Yeah. Okay. So, I'm kind of in this category, the framework builder, thought leader kind of, uh, learning design. I mean, these are things I'm interested in. Um, but just seeing this list doesn't really help me say, "Okay, so how do I know which ones for me?" And I guess they all sound good, but how do I prioritize them? Uh, let me ask it. How do I prioritize this? The emoji. Okay. Rule of thumb: Prioritize the intersection, not the options. Okay, I like that. Does this compound? I don't know. Like, already I feel not so energized looking at this list. I mean, I get it. It is very generic, like this is best practice, so to speak. But at the same time, you know, is it could also just be I glaze over everything. So, you know, just as a demonstration, I think it's going to give some good ideas. I mean, in the end, I do mostly three and four and a little bit of five as well, but, um, not so easy to get from this list to finding what I want to do. Okay, so that is the, you know, lack of context example.
Now, we can see with Claude Code, what does it look like when I actually share my thoughts unfiltered, straight in my Obsidian vault? Let me show you. All right. I'm using Claude Code here. It looks pretty intimidating, but it's actually not that difficult to set up. Anyway, um, I can share the instruction with you maybe in a different video, but the best place is actually Claude has a website. I'll link it right here that you can go there and it's quite easy to set up. All right. So, what I'm going to do is go into my Obsidian vault and help me understand what exactly am I interested in with creativity and framework thinking. Help me understand what is that intersection. Okay, I'm going to find the link. One second. See, it'll take some time. So, >> a few moments later, >> what does it find? I am fundamentally interested in building the case that creativity and structured thinking are not opposites. They are partners. My notes consistently reject the false dichotomy of logical and creative thinkers. Okay, it just articulated that better than me, 'cause I was thinking about this. I was like, "Okay, yeah, I know it's a paradox that I'm interested in." But actually, I've said this many times in my notes, and the key ideas or constraints enable creativity, and that creativity is a refined, refined mental models, and I really believe it's a proxy for intelligence. So, it's captured that word note. Uh, framework thinking is creative. I haven't really talked about this on the channel that much because I feel like, "Oh, you know, are people interested in this?" I really think creativity is at the top of Maslow's hierarchy, but I know a lot of us, you know, analytical people, we're like, "Creativity? Who's that for?" I think we're not quite understanding the word, and I have a problem with the word itself. But, you know, it's really the polymathy connection, right? Creative thinkers are often highly logical, right? Like Einstein was a great violinist. They did some study where most of the Nobel laureates had a very creative part of their life that they dedicate a good chunk of their time and energy to. And cross-disciplinary thinking really is a core at what I do. So, all right, what I seem to be building towards is a unified view where frameworks are the tools that break down complexity, creativity reassembling pieces into new holes is interesting. I'm interesting. I like this analysis and synthesis contrast. I didn't put that thought in anywhere explicitly in my vault. So, just having it go through, well, I went through 39,000 tokens, but having it going through all my notes and being able to just extract this is gold, right? My incognito conversation is never going to get me to this level. Uh, would I like to dig deeper into any specific notes? So, um, can you list down what parts of creativity I'm actually interested in? If I had a research, uh, question, what would that be? Give me a few options. Let's see what it'll say. I'll let it think for a bit. >> A few moments later. >> Okay, here we go. Uh, based on my notes, logic, creativity, unity, rea. Yeah. So, you know, rejecting the left brain, right brain. Well, I mean, I don't reject it, but I do think that both work together, and we all have a left brain and right brain. Um, okay. Constraint as creative fuel, structure input. I feel like this is already pretty overdone. So that's probably not yet, I think, is the, the, so what of that, and being able to explain this for those of us who are analytical and to understand that, "Okay, constraint, the, the, so what of having constraint is then we can be creative." Uh, okay. Intense learning, refining models and creative output. Oh, interesting. Again, I did not put this specifically in, but now that I see it articulated this way, it's giving me some ideas. This is, you know, this is the context that can get you what you are working on. Like, I'm like a kid in the candy store. Even though I, these are all my notes, but I wasn't going to go through each note one by one and come up with this, right? What I can say is, "Okay, here's everything I'm interested in that I've documented in my own words. Now, can you summarize it, synthesize it for me?" And that's an LLM strength. So, this is, I'm, I'm just a very happy girl right now. Okay. Cross-disciplinary thinking. I think I will, what I will do next is push it further. Okay. Actually, I'll do it with you here. Okay. So based on these, can you suggest some sources, resources to go deeper on books, uh, or authors, schools of thought so that I can build this out? Also, consider my other interests, um, put it in my vault so that I'm exploring the intersections of my interests. Okay, let's see how it goes. >> A few moments later. All right. So, five minutes later, here is what it shared. I already recognize a lot of these books, but, uh, I mean, it's good to know that they're aligned. I wonder if it saw my book list. I assume it did, but I like that it broke down some of the authors to follow, and, you know, I already follow them in this case. Okay. Process philosophy. Oh, I didn't know Whitehead. I do have Whitehead's book on creativity, but interesting process philosophy. I have to look into it. Um, I feel like the recommendations now kind of, um, you know, it's averaging down to what's very common. So I have to probably push a little bit here more. Maybe in Claude Code, maybe I will go to Elicit and look at more academic research or use research mode and get some of these as well. But, uh, here is interesting. But these, these are good. I mean, definitely know some of these the recommendations. I've read them. But, um, all right, that's good to know. I would now come back to the question of the career. Right. Okay. So knowing, you know, about my interests and questions for creativity and frameworks, help me, uh, design my career path. I want something that is practical and pushes forward our understanding of how to do knowledge work meaningfully and get paid for it. Okay, let's say something like this and see what kind of recommendation it would come up with. >> A few moments later. >> Okay, here we go. Um, it's mentioned. Okay, so first of all, the positioning. It knows what I'm doing. I'm glad this, uh, showed up as NA. Great. Um, the gap you could fill, productivity influencer. There is a missing voice in the knowledge work conversation, tactics without theory, academic theory without practice. Yeah, these are great. The gap is someone who can articulate how knowledge work actually works and translate it into teachable, applicable methods. Honestly, this is how I feel. I mean, I never really mentioned this in my notes, so it's kind of crazy that, uh, it articulated this. Building intellectual property for writing. I am writing a book currently. I don't know about the book to speak into corporate workshop to advisory type revenue stream. Yes, I know these are awesome people do it. Okay. The educator builder. Uh, okay. Okay. The applied researcher, work inside AI companies. I don't really want to do that. Hybrid phase one. Continue content but shift towards original frameworks, not just explaining others. Then productize. I don't know. I think this framework is very boring and actually I need to find some good examples of people I find interesting and actually let me do that. I find, okay, before I continue, I just realized now you're just indulging me in my me figuring out my future plans. So, if you're not interested in that, honestly, you can stop right here because, uh, I'm just going to continue doing this, uh, for the next little bit. But the point is, bring me back for having a note-taking system, right? What really continued to surprise me is how good LLMs are at taking information and being able to pull out some really great ideas and being able to identify, especially with connected notes, the core notes that matter to you. So, what it's able to do is without me needing to review all of my notes, it's able to pick out things like this. So if you are working on something and you want to build on top of it, you want the ideas to compound and not just build sandcastles, then you have to document it. It has to be explicit so that AI can work with it and help you see the things that you, as a human being, have, you know, natural weaknesses like digesting a lot of information in a short amount of time, right? So being able to engage in that and have your curiosity and your questioning mind and your judgment mind and your editorial mind come instead of just say, "Hey, you know, give me something," and then now, "Oh, this is not that great, right? So let me move on," and have that task sit in its current state is going to be so much more different than if you had these contexts so you can go deeper.
So with that, I hope you enjoyed this video. If you have any questions, let me know down below. I will link, um, the Claude Code setup information down there as well. All right, I'll see you in the next video.