Поиск по транскрипциям

elephant in the room. How does this compare to other AI coding tools? Because there's a lot of competition out there. You've got GitHub Copilot, Claude Code, Google AI Studio, Cursor. All these companies are fighting for the same space. Here's what makes Codeex different. Most AI coding tools are glorified autocomplete. You type something and they

right so essentially claude is just a chat in the cloud right so it's like chat GBT just a user interface like a simple chatbot then you've got claw code is fantastic for working offline um sorry for working locally for coding for building out like uh complex projects and for actually deploying like repositories from GitHub...locally, right? So, it can interact with your local files, it can build out projects, it can organize your desktop, that sort of thing. It's kind of like Claude code but for normal people, right? Normal nontechnical people. So, for example, if you go to Claude like so um you will get a version of co-work and then

click on the link in description right down below to get the source code from GitHub You are gonna wanna bring that into your IDE or I guess your Claude code if you're gonna go that way. First thing you wanna do is go and set up your environment variables. So after you've done your install

other day and you'll see this on one of my videos, I ran exactly the same prompt inside Claude as chat GPT and Claude outcoded it, right? Claude actually created an output. Chat GPT totally failed. And so the lesson there was like this AI is maybe good on the benchmarks, but in reality when you're using...than 50 LMS. " Yeah, I heard Amazon Bedrock is pretty good. Actually, the new Claude update for the million tokens context window that was only available in two places. Claude Code and Amazon Bedrock. It wasn't available anywhere else. That dude says, "What LM will allow us to schedule an agent to dedicate a task? " I can't remember

existing code that we have to make those changes that specified about improving the UI. Right? Some people ask like how do you have multiple agents modifying the code and that sort of thing. So number one you've got the agent on the right hand side and you can switch between like fast and planning to have these agents...runs in the background. And that was what was testing this out previously, right? The planner, the builder, etc. runs automatically over here. And then if you want the Claude coding agent to help you as well, then you just go to terminal, new terminal, run Claude. Boom. You're good to go. Okay. And now it's editing

developers and coding tasks. The fact that they're creating a specialized coding model tells us that XAI is serious about competing with tools like GitHub Copilot and Claude for coding. Here's something that really caught my attention. The benchmark leak shows that Gro 4's reasoning capabilities can be pushed even further with extra compute. The base model...like IMA scientific reasoning tests like GPQA and coding tests like SWE. It's hard to game that many different benchmarks simultaneously. The timing of this release is also crucial. OpenAI is rumored to be working on GPT5. Google is developing new versions of Gemini. Anthropic is pushing forward with Claude. The AI race is heating up and XAI needs

parallel tool calls. What does that mean? Claude Opus 45 can spin up multiple sub aents that all use multiple tool calls each. That means you can give Claude code a task and it can do research on the internet, create a document, look at different code files, write code and do all those at the exact same time. Then...many tool calls as it can. Do it all simultaneously. And what that's going to do is save you tons of time. So instead of waiting for Claude to use tool calls one by one by one, it'll say, "Hey, when you need to feel free to spin up sub agents and have them spin up tool calls

then you're going to go over to open to the API provider and select VSS code LM API and inside there you get free access to code with 03 mini Claude 3. 5 Sonic and all these other normally paid apis right so you can see them all there so you can code with all of those for free

then you're going to go over to open to the API provider and select VSS code LM API and inside there you get free access to code with 03 mini Claude 3. 5 Sonic and all these other normally paid apis right so you can see them all there so you can code with all of those for free...using virtual Studio code now inside the AI profit boardroom as well we actually have a section on how to set that up so if you go down the left hand side right here and you go to get co-pilot agent mode you find the sap on exactly how to set that up step by step because there

here's what just happened. Alibaba just dropped something that has the entire coding world talking. And I mean, everyone, senior developers, AI researchers, even the guys building Claude are probably looking over their shoulders right now. It's called Quen 3Coder 480BA 35B Instruct. I know that's a mouthful, but trust me, you need to know about this...paying anyone a single penny. But before I tell you exactly how good this thing is, let me ask you something. How much are you spending on coding tools right now? Claude Pro, ChatGpt Plus, Cursor, GitHub Copilot. Add it all up. I'll wait. Now, imagine getting something better than all of them combined for free. That's what

these would be claw code. We talked about it on the channel before if you're not familiar. This is a coding agent that runs inside of the command line. So inside of a terminal you simply say claude. When you install this I have multiple tutorial videos on this on the channel if you want to learn...claw code growing and people starting to experiment in this way. I mean it was nowhere close to as capable as it is now 6 months ago. And claude code literally came out end of February. So that's less than a year ago. So this is the type of thing that I'll be keeping

writing code, and when I wake up in the morning, the code will be fully done for me to review and ship to my app. Before when using claude code, you needed to be in your ID, right? sitting at your computer fully engaged looking at your monitor, sitting here writing prompts and wait for the agent to work. That

models context window. So you can have super long coding sessions without hitting limits. Let me talk about model support for a second. Open code works with over 75 different AI providers. Claude Sonet 4, GPT 4. 1, Gemini 2. 5, whatever you want. And you can switch between them midsession. If Claude is being slow, switch to Gemini...specific strengths, switch to that. Total flexibility. And here's what's really cool about this. Different AI models are good at different things. Claude is amazing at understanding complex context and writing clean code. GPT is great at creative solutions and edge case handling. Gemini is super fast and good at rapid iteration. With open code

example, before we went on to Claude, we said, "Okay, create a landing page for Julian Gold's SEO agency. " And we'll get the code from Claude and then we'll plug that into your wear. The other thing we can do as well is we can take the new deployment and then revamp that again. All right...after boost over here, we could take the code from that page, redeploy it, and then revamp it. So let's see what we get right here from Claude. But yeah, the main thing here is if you're a vibe coder or if you're learning how to code and that sort of thing, you can generate code

coding works. is moving from I write every line of code to I direct AI agents to build what I need. And tools like anti-gravity and claude code are leading this shift. By 2026, this is going to be normal. Everyone will be using AI agents to code. The people who learn these tools now, they're going

hello to the fly. Don't communicate with the fly. The fly is not welcome here. Do not communicate with the fly. They use your data to improve claude code so they can then go to corporations and sell them $10 million contracts for cloud code, which I get. That's smart. That's it's a business

stack here. There's the like action part of the stack, which is skills, sub aents, MCPS or APIs. again kind of fancy terminology for basically like running some code to do something. Uh those are going to actually interact with this concept of tasks which we're going to also look at in the same obsidian sort of flow...this is my morning plan skill and I've just vibe coded it. Like this is just a markdown file that's like a set of instructions that claude code can follow based on sort of its decision criteria if it wants to open up you know a scale or not. And so it says check new meetings in granola

there were some things I did not like with claude 3. 5 with codeex so or not codeex so hard to keep track of this claude code anyways um so we'll see what I think but anyways so this is a guey and basically uh what I want to do is just show you some

absolutely amazing. Can you please tell us more about Vibe Coding Academy? I like to join it. Well, the link's down below. You get a full Claude Code course. XGrowth Master Class. You get weekly calls with me. There's two calls of me this week. You get to ask me questions whenever you want. You get to talk

really boring, really mundane. - For me personally, since we're talking about coding, and you mentioned debugging... the source of enjoyment for me, more on the Cursor side than Claude Code, is that I have a friend, I have a pair programmer. It's less lonely. You made debugging sound like this great joy. No, I would say debugging