Transcription
Everyone is talking about ChatGPT, but for some reason, many ignore a tool like Claude. And, friends, I, uh, for a long time, I myself used ChatGPT. This tool was my most practical one. It contained all my projects. Perhaps you remember my old videos. I always championed ChatGPT. But lately, in the last few months, all my development projects have completely transitioned to Claude. And in a series of videos on my channel, I will fully reveal the Claude ecosystem, talk about all the practical features it has, everything related to writing code, and creating automation in N8N. We will reveal all of this, and on my channel, I want to fully reveal the amazing capabilities of Claude. So, today's lesson is a mini-express course, an introductory course, let's say, on this tool. Today we will talk about things like how to register here, how to create an account. This, I think, will also be useful for beginners. I will tell you about tools like artifacts and projects. I will also tell you what to click to make everything work great, how to interact with it in general. There is also a very subtle point that ChatGPT has, uh, its own style of interaction. Claude's is a bit different, it differs significantly. We will also talk about this in today's video. Therefore, friends, my name is Roman. You are on the Sukhov Life channel. Let's start getting acquainted with such a wonderful tool as Claude. Let's go. The first thing I want to draw your attention to is that it is impossible to access this tool from Russia. Like many advanced tools, such as ChatGPT, and probably all cool neural networks, it should only be used with a service from third-party books. I hope you have it. Therefore, as soon as you have it set up or installed, you go to the cloud.ai website. Also, for registration, friends, do not use an email address with a .ru domain. Use, of course, Google. For this, we simply click on "Continue with Google" and log in with one of your emails. Click "Continue." Next, we agree to all the rules that Anthropic offers us. Click "Continue." Here it asks for our date of birth. We register, enter our date of birth accordingly, and click "Continue." So, friends, regarding the plan, of course, I recommend you immediately get the Pro subscription, at least, because, uh, on the free account, you will have a small number of credits. They are limited here. If Pro has certain limitations, then in the free tariff, in the free one, there are practically none at all. But still, we will use this tariff today. At the end of the video, I will, of course, talk about the Pro tariff and its features, but we will choose the free tariff today, so follow along with me. What else? After you have chosen the tariff, Claude immediately offers to download the application for MacOS. Accordingly, you can download it, and the download starts immediately after you click on this icon. And you can already use Claude not only in the browser version but also install the application. The application must be downloaded because in subsequent lessons, we will use the application, not the browser version, but the application. Why will we need the application, you might ask, why not use the browser? In principle, it's the same, but there is still a difference. If we use a concept like MCP servers, we will cover this in future videos, then all interactions go through the application. Okay, we have downloaded it. I think you did it with me. Next, we can skip this part. And here we have a welcome message. "Hello, I'm Claude." It says here, if you have any questions, you can ask me. That is, you can ask me about any questions, and I will help you. It also says here that some chats and some answers from the neural network will be used by Anthropic for further training of its ecosystem. That is, they use conversations in chats with users to then train this model, to make it better and better. That is, they use some answers later to improve this model. But if you don't like something, you can, of course, change it in the settings. So I understand this. Understand? Before we get started, show you. So here it asks, how should I call you? That is, introduce yourself. Let's write Roman, and send it. Here it says, what will you use Claude for? That is, what are your goals? You can choose several, I think. Or no, a maximum of three. Well, let's say business strategy, content creation, coding, and so on. Well, for someone, maybe design is also important. Next, it offers to install Claude Code. Let's deal with Claude Code in future materials. You can, of course, click "Install." If you click "Install," you will go to the documentation. This is already a slightly more complex matter, because Claude Code cannot be deployed just like that. You can work with Claude Code, for example, within the terminal or within certain tools like Visual Studio Code. But we won't touch this topic for now; we will analyze the Claude neural network itself. Next, it asks where to start. Like with development. Create customer journeys. Well, it doesn't matter what we click here; in principle, any task is already interesting. So, it's loading. Friends, I don't know how it will turn out now. Everything is happening in test and real-time mode. In some cases, it may ask you for a phone number for registration, and in some cases, it may skip it directly. So, bam. And we already have some pre-prepared prompt, and it will do something here for us. Here the neural network has already helped us through a certain prompt to start some development. This is a kind of assistance for users so that they can interact directly with this neural network right away. Well, friends, I congratulate you. If everything worked out for you just like me, and you registered with Claude without any problems, now is the time to figure out what's here. We will talk about all these tools that are here. We will analyze them in more detail because if you are a beginner and opened Claude for the first time, there is a difference, especially for those who have used ChatGPT before. There is a significant difference here. Even if the interface looks similar at first glance, there is a place where we, for example, write a message, and where we receive it. These interfaces are essentially the same in all neural networks. But if we talk about the tools that are here, specifically projects, artifacts, code, and everything else, and also the tools that are here in this panel, where we type the text itself, they differ significantly, even when comparing with ChatGPT. First of all, friends, when we click on "New Chat," a new chat opens accordingly. Here you can, accordingly, write some query to the neural network. And what do we see here? The first thing to pay attention to on the free plan, on the free tariff, is the limited list of models available here. For example, on the free plan, we do not have the ability to communicate with such a cool, latest, cutting-edge model, Opus 4.6. We can only communicate with Sonnet 4.5. Regarding Sonnet 4.5, friends, this model is suitable for absolutely all tasks. I would probably classify it more towards tasks related to text. Everything related to text, rewriting, writing text, Sonnet is suitable for these tasks. Best everyday tasks, if we put it that way. If you want to engage in development, if you need to dive into more in-depth analysis, everything related to, for example, code development or automation, then please, Opus 4.6 is like an experienced developer who is, in principle, in your browser, and you can communicate with him regarding any development or automation creation. If we take a model like Haiku 4.5, it's the simplest model. It's quite dumb, so you can write something to it, and it will respond. On the free account, there are quite a lot of limits if you switch to Haiku. Well, and accordingly, there are other models, they are not available to us now; they are available on other tariff plans that are here. I will explain the difference between the free plan and, for example, the Pro plan a little later. So that you have some understanding of what the difference is. Okay, we have figured out this panel, right? I've told you everything about chats. So, what's convenient here, right? We can collapse this panel. That is, the windows with all these possibilities can be collapsed to fully focus on a task. This is very convenient. What else is here? A very good, convenient search among all chats. That is, when you click on "Search," for example, you type a specific word, and you can find a particular chat where you discussed a particular task. This is very convenient. This feature also exists in ChatGPT. It is also present in Claude. Regarding artifacts, we will also analyze them a little later with a practical example. This is a very cool thing when you can work within a single project, upload all the materials there, and Claude will understand the context and provide you with a specific result based on these materials. Artifacts are also a cool thing. It's related to the fact that within the browser version, you can do some development, some features, we will test this today as well. I really like this feature, especially when working with clients. They need to embed some mini-tool, a mini-application, or create a calculator for a website without any problems. Artifact handles this. You simply write the task, and Artifact, so to speak, deploys a specific development. And here is the code that recently appeared. Here, friends, it's a slightly more advanced story. We will analyze this a bit later in the next videos. Today, as part of getting acquainted with Claude itself, I just want to show what cool and interesting things are here. Today, we will simply analyze all these tools that are inside Claude. So, let's try to ask some kind of query. I am recording a lesson on N8N for beginners. Create a structure for 8-10 minutes. Hook, three key ideas, demo scenarios, typical mistakes, final conclusion, and so on. So, for example, we want to create a lesson structure. And what interests us? If you remember, ChatGPT has a lot of different models, and you can get confused, but here everything is quite simple. Well, for such a task, let's try to use at least Sonnet 4.5. Okay, we ask this query. It will be processed accordingly, and we will get an answer. We will get an answer. And now the neural network will give us the lesson structure for N8N. Showing on the screen, opening email, 50 new emails, switching to Notion, saying: "Now I will spend 2 hours copying and pasting between these services." But the key idea is that you can, of course, enable automation. And imagine such a demo scenario. It's still writing it. Typical mistakes, final conclusion. So, essentially, it has outlined such a task for me. As you can see, there is no context here, right? I said: "Create a structure for 8-10 minutes." That is, there is no understanding of what the lesson will be about, what will be used in this automation. It simply chose a specific topic from its head. The main thing I want to draw your attention to is that if you start a dialogue within a specific chat, it will be displayed in Russian here. Previously, by the way, it was like this: no matter what chat you started, it was always only in English. Now Russian is available here. Okay. If you want to create a new chat, you click "New," this chat remains here accordingly. And you can already, knowing the context, so to speak, if I need to continue, for example, this whole story within the lesson structure for N8N, I can continue to upload some materials here and communicate with Claude here specifically on this part. Well, when we start a new chat, a new chat is a new story. Accordingly, if we need a specific model, if we want to, for example, now we have a Pro tariff, and we want to use, for example, working with Claude, or creating some automation, we would use Opus 4.6. Well, we will work within Sonnet 4.5 for now. What else is interesting here? When we click on the plus sign, a pop-up menu appears. What can we do? Naturally, we can add any files or photos, add a screenshot. "Add project" is when we create a project or can start a new project right from here. We will get to that. "Web search" is a very good thing, it can also be turned off. This is searching for information on the internet, accordingly. "Style," this is a very cool thing. We can make the neural network communicate with us in a specific style, for example, in the style of a caring teacher, or in the style of a programmer, or for learning, or a corporate style, or simply a formal style, or you can set this style yourself by clicking here "Create," and here "Create custom style," for example, we can write an example style, for example, or upload a specific file, or paste text. Here we can describe the style in which we want the neural network to communicate with us. That is, you can make it as corporate as possible, so that it communicates with you like a friend in some moments. There are such cases. I have acquaintances who used a similar communication style in their corporate matters, so that the neural network works in such a zoomer format. That is, this was a thing among my younger colleagues. Therefore, you can set your own communication style here. This is very cool, by the way. It suits everyone, however they prefer. I also experimented with this, but I personally prefer, friends, a normal, formal style of communication. Claude communicates excellently, very restrained, very, very well. That is, it never babbles. What I like, to jump ahead a bit. If we communicate with ChatGPT on various matters, we can constantly ask it something, ask again, ask it questions, then here, we must be very structured. That is, if we ask a query, we cannot divide our query into several parts. All questions must be packed into our query so that all questions are asked at once. Then we will save on tokens and get a structured answer from Claude. This is the difference. I have already jumped ahead a bit and explained the difference. Okay, let's now work on a specific example. I will now send something here, and we will work with, for example, practical files that can be attached here. Well, for example, I recently received a proposal from Tinkoff Bank to open some new banking product, and I want to know what it consists of. Therefore, we will write the following prompt. "You are a financial and legal analyst. Analyze the attached PDF about 'Kubyshka,' credit line, and loan, and provide an ultra-short summary on one screen." That is, rules, and so on. That is, in principle, you can write it in your own words. Claude will understand you in this regard as well. Therefore, let's now set it up and see how it responds to this request. Okay, we see that it is a revolving credit line, with a daily fee from 8 to 240 rubles per day. That is, it has fully described the features of this banking product. Red flags: exorbitant interest rate. If, probably, the credit is not repaid on time, so to speak, with this thing, you can get a really very high rate. Automatic debiting, daily fee increases. That is, essentially, not a very profitable product that Tinkoff Bank offered me. That is, essentially, friends, we have now tried to analyze one of these important moments when we receive certain documents, and we need to analyze them, understand what is in them, because sometimes there can be certain, perhaps, red flags, as they are called, red flags, which we might not even notice when familiarizing ourselves with a particular document. In this regard, neural networks help us excellently. And even the Sonnet 4.5 model is good at this and coped with such a task. Let's now analyze artifacts. What are artifacts, friends? That is, within the browser, you can create some cool applications. What can it be? It can be some kind of profitability calculator, a loan agreement calculator, perhaps even some game. Look, there are even some games here. Here, within artifacts, users themselves can post, for example, their developments. Let's take Forest Explorer. Let's open it and see what it consists of. Here it says "3D Forest 3DS." And here at the bottom it says "Keep learning," like "View full chat." When we click here, we can interact with this artifact within the chat itself. And as you can see, it says: "Create 3D artifact z allow me." That is, by creating, so to speak, a 3D forest space, and we can move the mouse around, and see the entire 3D forest space. This is such a primitive development. We can also use it. The main thing is to see what's interesting here. We can see the development itself and the code it consists of. So, here's how we can see anything. And the only thing is, everything is written with a regular prompt, by simply creating a forest artifact. And we got this interesting 3D forest model as a result, so to speak. Let's try to create our own artifact in real-time. We click on "Artifacts" and click "New Artifact." And here it immediately suggests: "Well, what do you want to create? Choose, for example, a website, create some project, document management, games, specific product tools. Well, start from scratch." Let's choose this option and start interacting. "Help me create something from scratch." So, "start from scratch" means, well, start the project from the very beginning. And here it suggests what you would like to create. Well, I came to it with a ready-made prompt, so I want it to create an ROI automation calculator, how much time we spend now, the cost per hour, how many tasks per week. I have a similar calculator on my website 24i. You can also do certain things there. I want to see how it handles this task here. Therefore, in real-time, we enter this query. And it creates the calculator excellently. And as you can see, it's starting to write code now, which is very cool. That is, we are not doing anything. Our hands can be busy with something else. We can go, so to speak, pour ourselves some tea, get some cookies, and so on, and relax. During this time, it will be doing this feature for us. It's very cool if you, for example, work in design, you need to create some SVG graphics, something else specific, and then transfer all of this, copy it, as a specific code, and bring it into a specific application or website. And now our task is to create, essentially, such a calculator. Let's see how it handles it. Oh, it handled it very quickly, most importantly. Let's see what came out of it, how functional this development really is. So, time per task. So, if a person spends time on one task, for example, let's set 60 minutes. 60 minutes, number of hours per week, 20 hours of savings, and the number of tasks per week. If it's 22 tasks, then 20 hours saved, cost of work, let's set, for example, 100 dollars. And what about after automation? That is, time per task is 5 minutes. Savings in time, monetary rate, efficiency priority, and detailed statistics. So, it didn't just create some mini-calculator; it also created a whole analytics system for us, based on which we can not just position ROI automation as a product, as a feature, on our website, but there is also post-automation, hourly savings, and a whole analytics system with detailed statistics. So, we can also publish this, for example, using a link. How else can this be done? You need to show the client, as a certain product. You have created some feature, and you need them to test it quickly, within a specific project. So, now you don't even need to open VS Code, VS Studio, and all that. We just need to use artifacts to show the client some specific feature. Therefore, we have created a link. We simply copy it. Open it in any other browser, for example. And great. Our draft version of the development has opened in a separate browser. We can calmly test it, show it to the client or your colleague with whom you are working on the project, that I have developed a solution of this kind. Let's try to implement it. Well, and then, of course, we can look at both the frontend and the backend, right? That is, what this whole development consists of, all this code that is written, we can copy it and, so to speak, use this code, integrate it into some solution. What else would I emphasize in artifacts? For example, we can communicate within the artifact's dialogue, that is, within the context of this development. And, for example, ask it to do something with this thing that it has created for us. Specifically, this profitability calculator. For example, we can ask, "Make a dark theme." Well, just so that it completely redesigns our development into a dark theme. Okay. What is it doing now? The main thing is that you understand that it is creating version 2. That is, we can switch between the first and second versions. So, version one is what we had initially, and here is the second version that it is creating now. Let's wait for it to create a dark theme for us. That is, it is essentially completely redrawing it. And so on infinitely. That is, we can improve our development within one chat, modernize it, add something to it, and each time we will have new versions. This reminds me of, for example, the new version of N8N 2.0. We also save projects there now, and we can switch between versions in the same way. Here, please, it has created a dark theme for us very well. In my opinion, a dark theme looks very good everywhere. By the way, I will show you later how to switch to a dark theme in Claude as well. And like this. Yes, by the way, we can switch between versions, choose the second version, the third. That is, each time we need to change something here, it will create the subsequent versions. Subsequent versions, we added some new improvements, features, something else, and these new versions will constantly appear here. Well, I think we have figured out artifacts. Let's move on to the topic of projects. This is also an interesting point. So, friends, regarding projects, let's go to "Projects" and create a new project. Here we will write, I have an idea for a project we can create. For example, an analyzer of N8N automation. It suggests, gives hints. Hints on ready-made.
Ready. We are preparing the workflow. So, let's create a project. So, friends, what do we have here? First, we have certain instructions here. Instructions on how it should communicate within this chat, right? That is, we are creating a specific role for this particular tool here. This story somewhat reminds me of creating GPT with ChatGPT, but here the system is still slightly different. So, what will we do within the Digital Makers club, where I teach everyone automation creation. Soon, a branch on no-code will start. I have already filmed almost 25 lessons on N8N, and all the most practical things are available here. But what do we have inside the club? We have over 300 ready-made workflows for N8N. There is already a GPT assistant, but I want to create something for myself within the framework of being able to understand the automations that are in this archive. There are over 300 different automations here. And what do I want to do? Well, I have one of the documents called "Workflow for Digital Makers Club Members." And here is a description of different workflows that are uploaded to that zip file. I will simply take these two and transfer them here into files. Well, and also, I have prepared instructions on how it should handle these files. Let's insert this instruction here now. And what does it say here? You are an assistant navigator for the Digital Makers Cement Workflow Library. Your only sources are attached files with the database. First, how to work, clarify two requests, if there is not enough context, and so on. That is, we are saving these instructions. And what is most interesting, friends, here you can attach different tables for your project, different technical specifications, different documents, correspondence with clients, anything you want. That is, I have only uploaded 6% of the available buffer volume within this project. That is, I can upload even more files. That is, if there are any other specific documents in your projects, you can upload them without any problems, until the very end. Therefore, now we can already ask within these documents to give it a specific request. Give me, um, give me a list of automations, for example, Notion. Notion. Let me clarify a couple of points to select the most relevant workflow. So, while you are choosing options, allow me to find all workflows in the database. So, I have all the workflows, 18 of them, please. So, and the Notion database assistant, so now it gives me all the automations that exist, the names of these automations, automations for working with the content team through Notion, and so on and so forth. That is, it is now giving me all these moments completely. So, there are over 300 workflows in the document. And it found all these workflows that I need. What exactly do you want to automate in Notion? What else can it do for me? Here is the next point. We can now, for example, choose filling a database through Telegram forms. And here it already starts to suggest your level in N8N, intermediate, advanced, or beginner. Well, let's say I'm a beginner, I'm just starting this journey. And what is most interesting here? It will now suggest that I create some automation. So. So, look, a short summary for filling Notion Telegram, three workflow levels, that is, processing emails. What is needed as input? Incoming emails. What do we get? So, in fact, now we, as beginner, right, automators, can start building automation in N8N, for example, choose one or another automation, say: "I'm interested, for example, in Telegram and a multi-format bot, can I start some path now?" Decompose it for me, right, for example, this automation. What do I need to do to make the correct connection in each node? So. And we can already, within the existing database, find, for example, this file, upload it to N8N, and start creating these automations. What else is here? It's still working, look. Let's probably stop it. What is here? One to cloud response notify. All right, ready to start. Choose a workflow and say if you need help with specific steps. So, it acts as a mentor for N8N. And in fact, this can be your N8N trainer. So, if anyone needs such help, if you are now mastering N8N and want to get 300 ready-made workflows and such a wonderful instruction, you can write in the comments under this video. We will agree, and I will send you these files without any problems if needed. So, friends, we have dealt with the projects. Now we can move on to the next stage. So, friends, let's now break down some prompting basics about how to communicate with neural networks in general. For absolute beginners, I want to remind you once again of very important things, such as context. For example, how not to write. For example, if we write something like: "Write me a post about marketing." A post about marketing, right. What kind of marketing? What is the text about, what marketing? Well, it's a request about nothing at all. It's clear that the neural network will write something for me now, right. And it writes: "Marketing in 2026 is not about sales, but about trust." So, well, it will write something. In principle, the post might turn out to be not bad, but as you understand, there is no context here at all. But if we also create a new chat and provide more context, even if it's not a very long prompt, but within the framework of "write me five ideas for promoting online courses, Python for beginners. Budget 0 rubles." So, we have context, we have an understanding of what to write about. And it will now give five ideas for promoting an online Python course for beginners. I think that this, well, will turn out better, because content marketing, Telegram, YouTube, it now offers shorts, TikTok, so free lead magnet, so here you can insert absolutely any idea, so what you yourself face daily, what you yourself work on. So, if you need to get a specific idea at the moment, of course, you can use it instead of an online Python course, insert your topic here. Maybe you want to start, indeed, recording your courses. And here are ideas on how to indeed promote this course idea for Python. This is cool. So, friends, let's look at another example. For example, we want to write a letter to someone. Let's imagine a situation where we need to write a letter to the landlord asking for permission to get a cat. And here we specify the context: I have been living in this apartment for 2 years, I always pay on time. The neighbor on the floor has a dog. And here we write: "The tone is polite, but confident, not pleading." So, the task for our Claude is very clearly and understandably described. Here you go. Good day. I am writing to you regarding the possibility of getting a cat in the rented apartment. And the coolest thing is, you can directly send the email via Gmail. So, we have a connector with our mail, and we can actually send an email to our landlord via email, which is great. I won't do it now. In any case, with the latest versions of Claude, there are now many such possibilities that it automatically suggests to us. For example, to create a connector with Gmail or in the form of a test, it offers answer options. This is a very cool feature. I regularly work with Claude in development now. And these things that it suggests, even within a test, reduce token consumption. We will get to this point now, to the point that we are now on the free plan. But if you switch to the pro plan and above, there are also limits there, and we will talk about them at the end of the video. Therefore, what Claude offers now, it offers solutions that reduce the number of tokens consumed per request. This is very convenient. And if we have dealt with such basic things as communicating with Claude and writing prompts and maintaining context. I think that if you want to dive into this as a beginner in more detail, that is, how to communicate with neural networks in general, then I think you will find relevant materials on YouTube. There is no problem with that. Let's now dive into the settings. There are also many interesting things here. Well, for example, the connectors themselves, right, so that you understand what can be connected. I tried to do it with Gamma. So, for example, the same Claude can see projects in Gamma AI, that is, Google Drive presentations, for example, right? So that it sees the files that you have uploaded there. Of course, mail. That is, you can give it a request: "Analyze my email, find emails, for example, from 2024." For example, you are looking for some emails, Google Calendar or GitHub. That is, you can immediately upload projects there. Copy Belitors, we will analyze this very cool feature called Skills within Claude a little later. This will be a separate lesson. I will dedicate a separate video to this. We will figure out what skills are. This is very interesting. This is a real breakthrough, in my opinion. Therefore, we will work on this further. As for billing, billing is already linking your accounts, cards, right, if you have them, if you are from Russia, you can get an electronic card, right, virtual, and upgrade your plan. So, regarding the account, here we can, without any problems, see our location, delete our account. And in General, we also have these things, like a very interesting function. So, it says here: "Get notified if Claude finishes some work." For example, if you have assigned it a specific task, and it doesn't turn on for some reason, please check your browser settings. In short, if you have enabled some task, for example, code generation, and it is working, and you should receive a notification, then this thing can be turned on. Here we can set some specific personal preferences for how it can communicate with us, that is, for example, ask some questions before doing some specific work. So, here are examples. Example. Mostly I write code in Python, right, as one of the options. So, what language do you use? So, this is a very detailed, personalized thing where you can write anything about yourself, about what Claude should know about you. So, in principle, such a thing exists in ChatGPT, but the main thing is that you know where it is. The most important thing for me is the dark theme. Friends, I always switch to the dark theme, so here you can, of course, independently switch to the theme that is convenient for you. In principle, during all the time we have been recording the video, I have gotten used to the light theme, but the dark theme somehow, I have even relaxed a little now. It somehow evokes more comfort and tranquility. White light still somehow hurts the eyes. Well, it depends on what is more convenient for everyone. Now, friends, we will dive into my other Pro account, and I will show some other features that are in Claude, and in case you want to upgrade your account. So, friends, we have now dived into my account, which is a paid Pro account. I have two of them now. We have dived into one of them. Here in settings, when you pay, for example, switch to the Pro account, you will have the opportunity to view your limits. They are located here in the usage section. As you can see, there are daily limits. So, when you ask some heavy requests, as a rule, if it generates, for example, a ready-made file for you, if you are just chatting with it, right, asking a question, getting an answer, then the limits are not consumed so quickly. About 25 requests on average per day you can ask it. There are also weekly limits. They move a little slower than, of course, daily ones. But if you hit, for example, 100%, then, for example, if I work hard today, then considering that today is Friday, my limit will reset only on Sunday at 2 AM. So, here, as I said at the beginning of the video, right, comparison with ChatGPT, where we can communicate with ChatGPT without stopping, ask it constantly some iterations, questions, constantly ask it something again in the moment. And we have no limits. Although I once hit the limits of ChatGPT when I was working. Even a message popped up that you, Roman, have been working for a very long time, you should rest. So, here these limits are present, and the difference in responses is felt. Speaking of Claude, I want to say that it is very precise in its wording. If it doesn't understand something, it will ask you a question, it will ask you again. For example, Roman, what do you mean by this? Give me more context. And it will specifically ask you several questions so that you give it complete answers, and then it will give you the highest quality answer. This is where Claude's advantage lies. This is its internal intelligence, its clear restraint, and at the same time the quality of its answers and the files it provides. Whether you work with automation, whether you work with generating technical specifications, or some document flow. Claude does all this very clearly and competently, and most importantly, structured. There are no problems in this regard. That is, if you ask: "Create some document based on the analysis of some specific data," then it will do it no worse than, for example, Notalbook LM from Gemini. Therefore, this tool is very versatile, very cool. And everything related to, maybe at first it will seem to you that you are hitting these limits. Maybe it will even start to annoy you to some extent. But over time, when you work with Claude longer and more often, you will understand how to ask it questions correctly. And, in principle, this session, not the daily session limit, will be enough for you to solve almost all your tasks. I switched, once a month, to a more advanced plan. Let's now look at what plans are available if we want to upgrade. So, friends, here is the Max tariff, it offers higher limits. Priority queues. Yes, indeed, it happens that the load on Claude and some requests are not processed in the right way. Therefore, if anything, $100 is quite enough to code, right, create some projects, so that it works, in principle, 24/7 and creates some large-scale projects for you. It can work on several projects at once, which is very convenient. And I believe that if you really want to delve into development and use such a cool tool as Claude, in principle, $100 to pay for the tariff and do it qualitatively and coolly with such a cool assistant as Claude is not such high money. Therefore, when I have to work on several projects myself, right, and I need to have maximum focus everywhere, high quality and cool, then I, of course, switch to the Max tariff plan. That is, when I switch to this tariff, I know for sure that I have a lot of work and I need to do it qualitatively and on time. So, friends, thank you very much for supporting my channel, leaving comments, and providing activity under all my videos. This truly helps me to create even more quality content for you, to share wholeheartedly all the knowledge and skills that I acquire over time. Therefore, in the future, of course, there will be even more videos about Claude. I have now specifically created a whole playlist on Claude. We will analyze this tool in more detail, thoroughly, and each video will be practical, including so that you can use the power of this development for your own benefit. Therefore, friends, thank you very much for your attention. This was Roman, this is the Sukhov Life channel. See you in the next video. Goodbye everyone.