WEBVTT 00:00:00.000 --> 00:00:02.500 You know, it strikes me that we collectively 00:00:02.500 --> 00:00:05.759 misunderstand the tool we're holding. We look 00:00:05.759 --> 00:00:08.720 at ChatGPT, and we treat it like a poster. You 00:00:08.720 --> 00:00:11.179 put bread in, you get toast out. It's a transactional, 00:00:11.320 --> 00:00:14.099 single -use interaction. But looking at the research 00:00:14.099 --> 00:00:17.219 we have today, that is fundamentally the wrong 00:00:17.219 --> 00:00:20.199 mental model. Completely wrong. The real power 00:00:20.199 --> 00:00:23.120 isn't in the one -off question. It's in treating 00:00:23.120 --> 00:00:26.280 the AI like a system. Yes. A system that remembers, 00:00:26.460 --> 00:00:29.179 a system that learns, and a system that actually 00:00:29.179 --> 00:00:32.060 organizes your chaotic thoughts into something 00:00:32.060 --> 00:00:34.640 coherent. That's the disconnect right there. 00:00:34.820 --> 00:00:37.020 We treat it like a magic eight ball. Yeah. Shake 00:00:37.020 --> 00:00:39.960 it, get an answer, put it down. But the source 00:00:39.960 --> 00:00:43.020 material we're diving into today, systematize 00:00:43.020 --> 00:00:47.780 your workflow. 15 advanced chat GPT hacks. It 00:00:47.780 --> 00:00:50.259 just flips that script. It does. The core argument 00:00:50.259 --> 00:00:53.200 here is simple. If you find yourself typing the 00:00:53.200 --> 00:00:56.009 same context twice, You have already failed. 00:00:56.429 --> 00:00:57.890 Welcome to the Deep Drive. It is good to have 00:00:57.890 --> 00:01:00.009 you here. Today, we are going to try and move 00:01:00.009 --> 00:01:02.390 past the viral tricks. Please! We aren't doing 00:01:02.390 --> 00:01:04.469 write me a poem about a cat in the style of Shakespeare. 00:01:04.670 --> 00:01:06.930 Well, laughing. No, we are not. We want to get 00:01:06.930 --> 00:01:09.430 into the mechanics of building what the source 00:01:09.430 --> 00:01:13.090 calls a personal workflow engine. We have a roadmap 00:01:13.090 --> 00:01:15.549 today that covers setting up memory, automation, 00:01:16.269 --> 00:01:19.170 project structures, and some truly advanced decision 00:01:19.170 --> 00:01:21.530 -making tools. This is the difference between 00:01:21.530 --> 00:01:24.079 playing with a toy and hiring a chief of staff. 00:01:25.200 --> 00:01:27.280 And honestly, it starts with the thing most of 00:01:27.280 --> 00:01:29.439 us get completely wrong. What's that? Identity. 00:01:29.620 --> 00:01:32.560 Right. The foundation. The source calls this 00:01:32.560 --> 00:01:35.939 teaching chat GPT your writing style. I think 00:01:35.939 --> 00:01:38.140 this resonates because, well, how often do we 00:01:38.140 --> 00:01:40.620 find ourselves typing, make this sound more professional, 00:01:40.879 --> 00:01:43.519 or make this shorter? Stop using the word Dell. 00:01:43.700 --> 00:01:46.239 Every single day, it's prompt drift. You spend 00:01:46.239 --> 00:01:49.340 half your energy fighting the AI's default voice, 00:01:49.379 --> 00:01:52.879 which is usually that overly cheerful corporate 00:01:52.879 --> 00:01:55.079 customer service robot. It is exhausting. But 00:01:55.079 --> 00:01:57.319 is it fixable? Because I feel like I fix it in 00:01:57.319 --> 00:01:59.459 one chat, and then I open a new one, and we are 00:01:59.459 --> 00:02:01.700 back to square one. That's exactly what the source 00:02:01.700 --> 00:02:03.959 addresses. It gives us a specific mechanism to 00:02:03.959 --> 00:02:08.259 fix this permanently. OK. And the key is, don't 00:02:08.259 --> 00:02:10.919 describe your style. Show it. Show, don't tell. 00:02:11.139 --> 00:02:14.219 You don't say, I write casually. You take an 00:02:14.219 --> 00:02:17.199 email. You actually sent. when you're proud of, 00:02:17.379 --> 00:02:20.659 paste it in, and use a very specific prompt to 00:02:20.659 --> 00:02:22.979 lock it in. Let's look at that prompt, because 00:02:22.979 --> 00:02:25.939 phrasing seems important. It is. You say, analyze 00:02:25.939 --> 00:02:28.680 this for tone, style, format, and reading level. 00:02:29.300 --> 00:02:32.740 You let the AI break you down. It effectively 00:02:32.740 --> 00:02:36.080 deconstructs your linguistic DNA. And then, this 00:02:36.080 --> 00:02:38.500 is the kicker, you tell it to store that memory 00:02:38.500 --> 00:02:41.060 for a specific context. Like, remember, this 00:02:41.060 --> 00:02:44.020 is how I write to my team. That specific context 00:02:44.020 --> 00:02:46.400 part seems critical. It's everything. Because 00:02:46.400 --> 00:02:48.360 the way I write a script for this deep dive is 00:02:48.360 --> 00:02:50.680 very different from how I write an email to a 00:02:50.680 --> 00:02:53.979 confused accountant. Precisely. The source emphasizes 00:02:53.979 --> 00:02:57.039 one style per context. You can have a client 00:02:57.039 --> 00:02:59.659 email persona, a Twitter thread persona, and 00:02:59.659 --> 00:03:02.219 a strategic memo persona. All stored. All stored 00:03:02.219 --> 00:03:04.520 in memory. So next time, you just draft an email 00:03:04.520 --> 00:03:06.960 to the team and it pulls the team mask off the 00:03:06.960 --> 00:03:09.840 shelf. No more manual correction. I have to admit, 00:03:10.020 --> 00:03:12.939 I... I still struggle with this. I find myself 00:03:12.939 --> 00:03:15.460 manually correcting the tone in almost every 00:03:15.460 --> 00:03:17.080 session because I haven't taken the five minutes 00:03:17.080 --> 00:03:20.000 to set this up. It feels, I don't know, almost 00:03:20.000 --> 00:03:22.520 lazy on my part not to do it. It's not just you. 00:03:22.979 --> 00:03:25.159 The source mentions that personalization settings 00:03:25.159 --> 00:03:27.400 are the features most people skip. They're hidden 00:03:27.400 --> 00:03:30.060 away. They're hidden in the settings menu, but 00:03:30.060 --> 00:03:33.460 they act as a global baseline. You can go in 00:03:33.460 --> 00:03:36.919 there and set a default about you. your role, 00:03:37.280 --> 00:03:39.759 your experience level, your preference for no 00:03:39.759 --> 00:03:42.699 fluff answers. So you're giving the AI a permanent 00:03:42.699 --> 00:03:45.180 dossier on who you are. Exactly. So you don't 00:03:45.180 --> 00:03:47.020 have to introduce yourself every time you walk 00:03:47.020 --> 00:03:48.840 into the room. So does this mean we stop treating 00:03:48.840 --> 00:03:51.939 it like a generic robot? Yes. It becomes a mirror 00:03:51.939 --> 00:03:54.780 of your specific communication style. It stops 00:03:54.780 --> 00:03:57.159 sounding like an AI and starts sounding like 00:03:57.159 --> 00:04:00.259 you on your best day. That is a compelling thought. 00:04:00.560 --> 00:04:02.219 But to make that work, you need a container, 00:04:02.419 --> 00:04:04.300 right? You can't just have all these different 00:04:04.300 --> 00:04:07.560 personas floating around in a void. Right. This 00:04:07.560 --> 00:04:09.560 brings us to the second pillar of our discussion. 00:04:10.340 --> 00:04:13.819 Structure. Specifically, projects. Projects are 00:04:13.819 --> 00:04:16.899 huge. If you're still using the default chat 00:04:16.899 --> 00:04:19.879 history, you know, that infinite scroll of random 00:04:19.879 --> 00:04:21.439 conversations on the left side of the screen. 00:04:21.540 --> 00:04:24.160 Which is most people. It's most people. You are 00:04:24.160 --> 00:04:27.040 living in a messy house. The source describes 00:04:27.040 --> 00:04:30.680 projects as folders with a brain. Folders with 00:04:30.680 --> 00:04:32.740 a brain. Break that down for me. Think about 00:04:32.740 --> 00:04:35.480 your physical desk. You wouldn't stack your tax 00:04:35.480 --> 00:04:37.800 returns, your creative writing journal, and your 00:04:37.800 --> 00:04:40.519 grocery list in one pile, would you? I mean, 00:04:40.620 --> 00:04:44.240 I might, but I shouldn't. Exactly. Right now, 00:04:44.500 --> 00:04:47.639 most people's chat history is that pile. Projects 00:04:47.639 --> 00:04:50.180 let you create distinct workspaces. Okay. One 00:04:50.180 --> 00:04:52.699 project for client work, one for creative writing, 00:04:53.100 --> 00:04:55.459 one for learning Python. And here's the key. 00:04:55.959 --> 00:04:58.139 Each project has its own custom instructions 00:04:58.139 --> 00:05:00.560 and its own files. And there is a specific setting 00:05:00.560 --> 00:05:02.680 here that the source flags as credible project 00:05:02.680 --> 00:05:06.120 -only memory. Yes. This is the safety lock. By 00:05:06.120 --> 00:05:09.920 default, chat GPT's memory can bleed across conversations. 00:05:10.360 --> 00:05:13.279 You do not want the casual, slang -filled tone 00:05:13.279 --> 00:05:15.740 you used in your creative writing project to 00:05:15.740 --> 00:05:17.680 suddenly show up. In your corporate strategy 00:05:17.680 --> 00:05:19.660 project. That would be a career -limiting move. 00:05:19.959 --> 00:05:22.879 It would be awkward, to say the least. Project 00:05:22.879 --> 00:05:26.259 -only memory isolates the context. It ensures 00:05:26.259 --> 00:05:28.699 that what happens in the project stays in the 00:05:28.699 --> 00:05:31.300 project. Right. It's essential for privacy too, 00:05:31.319 --> 00:05:33.199 especially if you're uploading sensitive client 00:05:33.199 --> 00:05:35.980 docs. It reminds me of the concept of context 00:05:35.980 --> 00:05:39.199 switching in human psychology. It takes energy 00:05:39.199 --> 00:05:42.480 to switch from one task to another. By creating 00:05:42.480 --> 00:05:45.000 these projects, you're effectively offloading 00:05:45.000 --> 00:05:47.500 that cognitive load onto the machine. That's 00:05:47.500 --> 00:05:49.600 the perfect analogy. You're building distinct 00:05:49.600 --> 00:05:51.920 rooms for distinct modes of thinking. That's 00:05:51.920 --> 00:05:54.100 it. And sometimes, you don't even want a room. 00:05:54.300 --> 00:05:56.879 Sometimes you just want a sandbox. The source 00:05:56.879 --> 00:05:59.399 mentions temporary chats for this. Right, the 00:05:59.399 --> 00:06:03.040 incognito mode of AI. Basically. These are chats 00:06:03.040 --> 00:06:06.139 that don't save, don't train the model, and don't 00:06:06.139 --> 00:06:08.680 create memories. Okay, so what's that for? It's 00:06:08.680 --> 00:06:10.839 perfect for when you need to ask a dumb question, 00:06:11.220 --> 00:06:13.959 or test a prompt, or run a sensitive role play, 00:06:14.079 --> 00:06:16.300 and you just want the slate wiped clean immediately 00:06:16.300 --> 00:06:19.019 after. It keeps your main memory banks from getting 00:06:19.019 --> 00:06:21.839 polluted with garbage. Why is the isolation of 00:06:21.839 --> 00:06:24.899 memory so critical here? It's context hygiene. 00:06:25.240 --> 00:06:27.240 You don't want your casual chats leaking into 00:06:27.240 --> 00:06:29.360 professional strategy. You keep the signal to 00:06:29.360 --> 00:06:32.560 noise ratio high. Cleanliness is next to godliness, 00:06:32.699 --> 00:06:37.339 even in code. Right. Now let's move to something 00:06:37.339 --> 00:06:39.519 that I think changes the dynamic of the relationship 00:06:39.519 --> 00:06:42.120 entirely. We're talking about automation. Yeah. 00:06:42.259 --> 00:06:45.399 The shift from passive to active. This is where 00:06:45.399 --> 00:06:48.769 it gets fun. Most of us think of Chat GPT as 00:06:48.769 --> 00:06:51.529 a tool we have to pick up and use, but the source 00:06:51.529 --> 00:06:53.509 points out that you can flip that dynamic. You 00:06:53.509 --> 00:06:56.129 can set up a daily reminder system. Give us an 00:06:56.129 --> 00:06:57.970 example of how that works practically because 00:06:57.970 --> 00:07:00.410 my phone already has reminders. True, but your 00:07:00.410 --> 00:07:03.670 phone just pings you with a label. Chat GPT can 00:07:03.670 --> 00:07:06.829 perform a task. A task. You can tell it. Every 00:07:06.829 --> 00:07:09.629 weekday at 8 a .m. send me three important AI 00:07:09.629 --> 00:07:13.410 updates focused on OpenAI and Google. It creates 00:07:13.410 --> 00:07:15.949 a background task. It runs whether you open the 00:07:15.949 --> 00:07:18.730 app or not. So it pushes the information to me. 00:07:18.810 --> 00:07:21.149 It pings you with the briefing. That is a subtle 00:07:21.149 --> 00:07:23.730 but profound shift. It's no longer waiting for 00:07:23.730 --> 00:07:27.230 input. It is initiating the interaction. It's 00:07:27.230 --> 00:07:29.430 massive. And you can scale this up with what 00:07:29.430 --> 00:07:31.569 the source calls the company knowledge base. 00:07:31.709 --> 00:07:34.509 Right, for teams. If you're on a team plan, you 00:07:34.509 --> 00:07:37.529 can connect your Google Drive, your Notion, your 00:07:37.529 --> 00:07:40.689 calendars. So instead of asking chat GPT to search 00:07:40.689 --> 00:07:43.439 the web, You're asking it, what did we decide 00:07:43.439 --> 00:07:45.899 in last week's marketing meeting? And it's pulling 00:07:45.899 --> 00:07:49.019 the answer from your actual internal doc. From 00:07:49.019 --> 00:07:51.319 your stuff. And there's also this mention of 00:07:51.319 --> 00:07:54.639 apps and connectors using the at symbol. I've 00:07:54.639 --> 00:07:56.500 seen that. Yeah. This is the app layer. You type 00:07:56.500 --> 00:07:59.259 at in the chat and you can call up external tools. 00:07:59.339 --> 00:08:01.839 You can summon Khan Academy for a math explanation 00:08:01.839 --> 00:08:05.199 or a design tool to generate a vector image right 00:08:05.199 --> 00:08:07.339 there in the flow of conversation. So no more 00:08:07.339 --> 00:08:10.490 tap switching. It's orchestration. It's fascinating. 00:08:10.769 --> 00:08:14.689 Imagine scaling this. Waking up to a personalized 00:08:14.689 --> 00:08:16.629 briefing created by an agent that ran through 00:08:16.629 --> 00:08:18.490 your internal docs and the external web while 00:08:18.490 --> 00:08:21.790 you slept. It feels like we are inching closer 00:08:21.790 --> 00:08:24.889 to that sci -fi vision of a digital butler. Ideally 00:08:24.889 --> 00:08:26.970 a butler that doesn't judge my calendar management. 00:08:27.689 --> 00:08:32.009 But yes, that's the dream. So if we look at this 00:08:32.009 --> 00:08:35.590 automation layer, it stops being passive. Right. 00:08:35.629 --> 00:08:38.909 It transitions from a tool you hold to an assistant 00:08:38.909 --> 00:08:41.700 that nudges you. It becomes proactive. I want 00:08:41.700 --> 00:08:43.519 to pause for a moment to consider the implications 00:08:43.519 --> 00:08:45.519 of that. But first, let's take a quick break. 00:08:45.860 --> 00:08:49.019 We are back. We've talked about identity, structure, 00:08:49.340 --> 00:08:51.840 and automation. Now I want to get into the act 00:08:51.840 --> 00:08:54.779 of creation itself. Yes. The source mentions 00:08:54.779 --> 00:08:57.159 advanced creation, specifically dealing with 00:08:57.159 --> 00:09:00.200 how we generate output images, code, mini apps. 00:09:00.419 --> 00:09:02.679 This is for the builders. The first hack here 00:09:02.679 --> 00:09:05.779 is model switching. inside custom GPTs. This 00:09:05.779 --> 00:09:07.919 is a bit technical, but super important. Let's 00:09:07.919 --> 00:09:10.039 break it down. People build these custom assistants, 00:09:10.659 --> 00:09:13.600 maybe a writing coach, GPT, but they built it 00:09:13.600 --> 00:09:16.019 six months ago. It's running on an old brain. 00:09:16.200 --> 00:09:18.360 And the source said you can swap that brain out. 00:09:18.600 --> 00:09:20.799 Exactly. You can go into the settings of an old 00:09:20.799 --> 00:09:23.399 GPT and force it to use the new thinking models 00:09:23.399 --> 00:09:25.679 like 01 for deep reasoning. It's like taking 00:09:25.679 --> 00:09:28.240 a vintage car and dropping a Ferrari engine inside. 00:09:28.379 --> 00:09:30.820 That's it. You keep the interface, but the performance 00:09:30.820 --> 00:09:34.259 skyrockets. Speaking of performance, the image 00:09:34.259 --> 00:09:36.639 generation workflow described here seems much 00:09:36.639 --> 00:09:39.379 faster than the usual method. It's the batch 00:09:39.379 --> 00:09:43.039 images hack. Usually people prompt, make a logo. 00:09:43.159 --> 00:09:47.509 Wait, make it blue. Wait. It's so slow. Yeah, 00:09:47.549 --> 00:09:50.269 it's iterative. The source says do it in parallel? 00:09:50.809 --> 00:09:53.850 Tell it. Create a hero image for a sauce website. 00:09:54.549 --> 00:09:57.830 Also create a dark mode version and a minimalist 00:09:57.830 --> 00:10:00.409 thumbnail version. So you get three variations 00:10:00.409 --> 00:10:02.970 instantly. And you can spot the direction you 00:10:02.970 --> 00:10:05.549 like immediately without waiting for three separate 00:10:05.549 --> 00:10:08.600 failures. That brings us to canvas mode. This 00:10:08.600 --> 00:10:10.679 is something I've seen pop up recently. It seems 00:10:10.679 --> 00:10:12.820 to be more than just a chat window, but less 00:10:12.820 --> 00:10:15.259 than a code editor. Canvas is incredible. It's 00:10:15.259 --> 00:10:17.259 essentially a prototyping environment. If you 00:10:17.259 --> 00:10:20.159 ask ChatGPT to build a to -do list app and you're 00:10:20.159 --> 00:10:22.659 in Canvas mode, it doesn't just write the code 00:10:22.659 --> 00:10:25.080 snippet. It renders a preview. You can click 00:10:25.080 --> 00:10:26.980 the buttons. You can check the boxes. So you 00:10:26.980 --> 00:10:29.480 are building software without writing a line 00:10:29.480 --> 00:10:31.740 of code. You're building interfaces. Calculators, 00:10:31.960 --> 00:10:33.980 planners, simple interactive tools. You describe 00:10:33.980 --> 00:10:36.460 it, it appears. What does Canvas mode actually 00:10:36.460 --> 00:10:39.519 represent for a non -coder? Instant prototyping. 00:10:39.580 --> 00:10:42.460 It turns an abstract idea into a clickable reality 00:10:42.460 --> 00:10:45.340 in seconds. It removes the fear of the blank 00:10:45.340 --> 00:10:47.679 page because you can see the thing working immediately. 00:10:48.179 --> 00:10:50.559 That removal of friction is the theme here, but 00:10:50.559 --> 00:10:52.860 I want to go deeper. Let's do it. The final section 00:10:52.860 --> 00:10:56.860 of our roadmap covers deep thinking. We are moving 00:10:56.860 --> 00:11:00.419 beyond simple tasks to research, strategy, and 00:11:00.419 --> 00:11:03.080 complex decision making. This is where the power 00:11:03.080 --> 00:11:06.879 users live. The source highlights deep research. 00:11:07.070 --> 00:11:10.190 combined with your own data. OK. Now, deep research 00:11:10.190 --> 00:11:12.549 is an agent that goes out, reads dozens of websites, 00:11:12.750 --> 00:11:15.129 and compiles a report. That's great. But the 00:11:15.129 --> 00:11:17.710 magic happens when you tell it to cross -reference 00:11:17.710 --> 00:11:20.490 that web data with your uploaded internal strategy 00:11:20.490 --> 00:11:23.009 documents. So it's validating your internal assumptions 00:11:23.009 --> 00:11:25.730 against the external reality of the market. Precisely. 00:11:25.909 --> 00:11:28.370 It's a reality check machine. And speaking of 00:11:28.370 --> 00:11:31.090 reality checks, there's this decision simulator. 00:11:31.490 --> 00:11:33.730 I love this one. How does that work? You're stuck 00:11:33.730 --> 00:11:36.860 on a big decision. Maybe, should we? pivot to 00:11:36.860 --> 00:11:39.620 enterprise clients. Instead of asking for advice, 00:11:39.759 --> 00:11:41.559 which can be vague, you ask for a simulation. 00:11:41.820 --> 00:11:44.659 A simulation. You say, simulate this decision 00:11:44.659 --> 00:11:47.879 from three perspectives. My customer, a competitor. 00:11:48.139 --> 00:11:50.740 and a skeptical investor. A skeptical investor. 00:11:50.879 --> 00:11:52.840 That's a useful perspective to have on demand. 00:11:53.039 --> 00:11:55.940 It is. And the AI role plays all three. It predicts 00:11:55.940 --> 00:11:58.279 the friction points, the counter moves by the 00:11:58.279 --> 00:12:00.740 competitor, the doubts of the investor. It helps 00:12:00.740 --> 00:12:02.620 you see your blind spots. It's not a crystal 00:12:02.620 --> 00:12:05.620 ball. No, but it clarifies the risk. That is 00:12:05.620 --> 00:12:08.500 powerful. There is also a tool mentioned for 00:12:08.500 --> 00:12:11.559 documentation, the SOP generator. But the method 00:12:11.559 --> 00:12:14.399 described is conversational. This is the interview 00:12:14.399 --> 00:12:16.830 mode hack. Yeah. Most people hate writing standard 00:12:16.830 --> 00:12:19.970 operating procedures, SOPs. It's boring. It's 00:12:19.970 --> 00:12:22.350 so boring. So instead of writing it, you tell 00:12:22.350 --> 00:12:25.149 chat GPT. Ask me clarifying questions one by 00:12:25.149 --> 00:12:27.470 one about how I do this task and then turn my 00:12:27.470 --> 00:12:29.830 answers into a detailed SOP. So it interviews 00:12:29.830 --> 00:12:32.649 you about your own job. Yes. You just brain dump 00:12:32.649 --> 00:12:35.509 the answers verbally or in quick text and it 00:12:35.509 --> 00:12:37.590 structures the chaos into a formal document. 00:12:37.730 --> 00:12:40.169 It captures your tacit knowledge. The stuff you 00:12:40.169 --> 00:12:42.350 do without thinking and writes it down for you. 00:12:42.509 --> 00:12:45.519 And finally, there is the concept of the knowledge 00:12:45.519 --> 00:12:48.500 graph. This feels like a long -term project. 00:12:48.840 --> 00:12:51.059 This is for the lifelong learners. Yeah. You 00:12:51.059 --> 00:12:53.019 create a project called Thinking and Research. 00:12:53.360 --> 00:12:55.559 You dump your book notes, your random ideas, 00:12:55.960 --> 00:12:58.620 your article summaries into it over months. Okay. 00:12:59.240 --> 00:13:01.879 And periodically you ask it, what patterns are 00:13:01.879 --> 00:13:04.440 emerging in my thinking or connect this new idea 00:13:04.440 --> 00:13:07.279 to things I wrote three months ago? It's looking 00:13:07.279 --> 00:13:09.460 for the connective tissue in your own intellect. 00:13:09.679 --> 00:13:12.220 It is. It's finding the threads you forgot you 00:13:12.220 --> 00:13:14.990 were weaving. What is the ultimate value of the 00:13:14.990 --> 00:13:18.190 knowledge graph approach? Synthesis. It finds 00:13:18.190 --> 00:13:20.210 connections between your ideas that you might 00:13:20.210 --> 00:13:22.570 have forgotten. It stops your knowledge from 00:13:22.570 --> 00:13:24.629 being just a pile of notes and turns it into 00:13:24.629 --> 00:13:28.070 a web of understanding. Synthesis. That is the 00:13:28.070 --> 00:13:30.350 goal. We have covered a lot of ground here. We 00:13:30.350 --> 00:13:33.250 have. From establishing an identity to building 00:13:33.250 --> 00:13:36.309 a structure, automating the mundane, prototyping 00:13:36.309 --> 00:13:39.590 ideas, and finally, using the machine to deepen 00:13:39.590 --> 00:13:42.340 our own thinking. It's a lot. And the source 00:13:42.340 --> 00:13:45.120 makes a great point in the big idea recap. The 00:13:45.120 --> 00:13:47.279 people getting the most value aren't the ones 00:13:47.279 --> 00:13:50.659 writing the perfect complex super prompt. No. 00:13:50.799 --> 00:13:52.820 They're the ones building systems. It is a shift 00:13:52.820 --> 00:13:56.580 from chat bot, which is transactional, to operating 00:13:56.580 --> 00:13:58.960 system, which is cumulative. Right. And the advice 00:13:58.960 --> 00:14:02.840 is don't try to do all 15 of these things tomorrow. 00:14:03.039 --> 00:14:05.940 You will burn out. Just pick one. Pick one. Maybe 00:14:05.940 --> 00:14:08.120 just set up the project only memory so your work 00:14:08.120 --> 00:14:11.080 stays clean. Or maybe just try the writing style 00:14:11.080 --> 00:14:13.440 hack so you stop fighting the tone. Stack them 00:14:13.440 --> 00:14:16.799 gradually? Exactly. Lego blocks, not a monolith. 00:14:17.480 --> 00:14:20.000 So here is our challenge to you, the listener. 00:14:20.659 --> 00:14:24.399 Pick one. Try the decision simulator on a choice 00:14:24.399 --> 00:14:27.139 you're avoiding making. Or try the writing style 00:14:27.139 --> 00:14:28.980 hack and see if it feels like looking in a mirror. 00:14:29.500 --> 00:14:32.279 And honestly, try the SOP generator if you hate 00:14:32.279 --> 00:14:35.419 documentation. It's a lifesaver. It really sounds 00:14:35.419 --> 00:14:38.019 like it. You know, as we close, I'm left with 00:14:38.019 --> 00:14:41.299 this thought about memory. We're outsourcing 00:14:41.299 --> 00:14:44.299 not just our tasks, but the retention of our 00:14:44.299 --> 00:14:48.120 context. If the AI remembers our style, our past 00:14:48.120 --> 00:14:49.960 projects, our patterns of thought better than 00:14:49.960 --> 00:14:52.980 we do, it changes how we view our own past work. 00:14:53.279 --> 00:14:55.379 It's no longer something we leave behind. It's 00:14:55.379 --> 00:14:57.639 something that actively informs what we do next. 00:14:57.879 --> 00:14:59.659 It makes everything we've ever done available 00:14:59.659 --> 00:15:03.269 to us right now. That's a superpower. A superpower 00:15:03.269 --> 00:15:05.549 indeed. Thank you for listening to this deep 00:15:05.549 --> 00:15:06.990 dive. Catch you on the next one.