just telling it to go and find all the different types of luxury restroom trailers that a business offers. And I do like to add to just like get Claude Code's opinion. Give me your game plan. Tell me if I'm missing anything. and just let me know before we go for it. — Before we burn tokens, tokens...time. And this is also helpful cuz after this pass, I could examine the results. You always find edge cases, just these one-off situations where you should let Claude Code know that it messed up. It done messed up so that you can fix it. And um you might have to rerun this. I reran this, I think
Поиск по транскрипциям
this new release of Codeex CLI that they, by the way, fully open sourced is an Agentic product that works in the terminal and does pretty much exactly what Claude Code does, which was released by Enthropic about 2 months ago, and I've been using that tool regularly since its release. It's really great. It basically runs...drop images into your terminal and let it build little applications for you on demand on full autopilot. It's obviously the same thing that Entropic did with claude code and I think it's amazing that OpenAI decided to open source this entire program and they said there will be more coming but as of now if there
model also getting better at these sorts of things? as we make the product features itself better. - It's pretty organic, honestly. Like, you know, at Anthropic, everyone uses Claude Code. And that includes the researchers. And so every day the people building the models are using the model in order to do their job. And I think as part...that you hit with a model. So, you know, as an example, maybe the model's really bad at doing certain kinds of edits. And sometimes when you use Claude Code, you see like, oh, failed to replace string, failed to replace string. Like, this is a model capability, and we can improve this if we learn from
Okay, now if you run Claude Code, one of the biggest changes you will see is that they're planning to deprecate output styles. So if you do `/output-style`, you will see that it will be removed on November 5th or later, and they'll automatically convert any output styles you have. And they basically recommend using plugins instead...system prompt files, system-prompt, append-system-prompt, or any of these, basically whichever one gets the job done easily. I do go through all of them in my Claude Code masterclass as well, so do check that
biggest of which is previews where Claude Code can run a website locally and then click through it to make sure that everything works as intended and take screenshots along the way. So let's quickly test it out. So clicking over here, you want to switch to local and you can see they added the work tree functionality...change this to a green button, press enter. And now this button is green. So this can be pretty handy for design-related work and also getting Claude Code to click around and behave like a user for you to make sure things work as intended. So I can also do persist sessions as well so that any cookies
some people in the fin fam. And we also ship. You do all the above. If you're just here just for the tutorials on AI and Claude code, I got a ton of main videos on the YouTube channel you can watch. This is our 2hour AI hangout session. Enjoy, get to meet people, find co-founders, learn...being agentic and doing it every day and doing it over and over again and learning as you go. Those are the two things that matter. Just watch your Claude code file system interview you did yesterday. I started implementing the same day in Obsidian. Great stuff. I didn't do that interview yesterday. I did that interview like
going to put GPT5 and Claude Code through the paces. We're going to put them through the world famous Alex Finn AI model benchmarking test. This benchmark test consists of four programming tests. We're going to build a 3D firstperson shooter, an Elon dancing animation, a 3D city flyth through, and a music visualizer. After we go through...these four intensive, mind-blowing world famous tests, we will know which of the models is better and if you should stick with Claude Code or if it's time to switch to GPT5 and the King's been dethroned. I'm also going to include all these prompts down below. So, if you want to take the world famous
think I've been at any other point in my career in AI space so far. when I see the productivity gains that I'm getting from things like claude code and building workspaces to be able to operate all my companies from a central place to be able to pull in data and make decisions and have a true...just build the most insane stuff just with a couple birds just recently the massive leaps in productivity I've been able to get out of setting up claude code workspaces which I've done a whole video on if you guys haven't watched that highly recommend it I'll put that up here
just go back and then I'll say, great. Do you have access to my ClickUp MCP? I'll say, awesome. Create a new content idea called Claude Code course. Now what it's going to do is scroll through all of my lists. It's then going to search for various ones that I may or may not have...going to ask it to insert it in my main YouTube queue. So I'm going to do that. And now you can see there's a task called Claude Code course. If I go back here to my ClickUp, open up the specific task. There's now a course that's basically been created. I don't like
give uh Claude two powerful tools the first is the stream native MCP server that coin talked about uh it exposes the full Pulsar admin API directly to Claude code so Claude can quickly query topic stats inspect schemas check function status and manage namespace policies through natural language without us having to write a single line of code...fall outside of the MCP service uh specifically uploading compiled packages to the registry. That's something MCP does not do yet. And between these two tool to tools, Claude could go from I have no idea what's wrong to root cause uh confirmed and actually I'll even show you solving the problem uh in just
give uh Claude two powerful tools. The first is the stream native MCP server that Kundan talked about. Uh it exposes the full Pulsar admin API directly to Claude code. So Claude can quickly query topic stats, inspect schemas, check function status, and manage namespace policies through natural language without us having to write a single line of code...fall outside of the MCP service uh specifically uploading compiled packages to the registry. That's something MCP does not do yet. And between these two tool to tools, Claude could go from I have no idea what's wrong to root cause uh confirmed and actually I'll even show you solving the problem uh in just
this puts Claude Sonnet 4 on par with Gemini 1. 5 Pro, which also has a 1 million token context window, and is quite a bit cheaper than Claude Code or Claude Sonnet 4 over here. Basically
just another AI chatbot. This is like having a junior developer, researcher, and content creator all rolled into one. Google just declared war on every other AI coding tool. Claude Code from Anthropic. Good luck competing with free. Open AI codeex. They charge you for everything. Google gives it away. GitHub Copilot. This does way more than just code completion
right now. First, install Node. js version 22 if you don't have it. Second, make sure your Chrome is updated. Third, pick an AI coding assistant. Claude Code is great. Cursor works well. Gemini CLI is solid. Fourth, install Chrome DevTools MCP. Run the MPX command or add it to your config. Fifth, test it on something simple
code? So, claude code is this like agent that you can use um in a command line interface. So, it's just basically this tool you can use that can control your computer and you can use it through natural language. Right? So, I can say make a file or make a file on my desktop that says hello Greg...create a new session to explain that all I don't want to have to explain that over and over again. Um, a lot of people are using like claude uh or chatbt on the web and it has things like memory but you can't like control you don't know what's in that memory right
coding, but here's Google DeepMind CEO Deis on those same scaling laws. — Scaling laws um are going very well. So we're definitely seeing increased capabilities by putting in more compute, more data, uh, and making these models generally larger. So that trend is continuing. Um, may not be not as fast as it was a couple of years...topic comes up. Just personally for me, the smartest thing to do is to not discount the possibility of a rapid takeoff of capabilities. Indeed, lean into tools like Claude Code and Claude Co-work so you can see just how good they are and the mistakes that they still make. But don't bet your future on that imminent
template, so can generative ai, today's tools, uh, particularly the newest, uh, versions, and we're gonna talk about this at Social Media Marketing World 2026, um, like Claude Code and Claude Cowork, which are agents, you give them templates, you give them recipes, and you come back later and it's done. I was just getting ready
step thinking you can see some examples of it right here and this could be absolutely insane on top of that they're also releasing an agentic coding tool Claude code we'll come on to that later
including like some of the best automations with AO3 and all sorts of stuff like this. So let's see how they perform. Now we've got Gemini coding out Claude's coding out as well. One thing we can do in the meantime is just compare the costs of each model as we wait for these tools to build...built out tool like you can see. Let's open this up. We have the results back. All right. So if we actually have a look at Claude's example, we've got more functionality. Let's just check it actually works. Yes, it does. Let's try rectangle. Yeah, look at that. So, it's actually created like
five months ago worth 183 [clears throat] billion now 380 billion second biggest private funding round in tech history after open AAI revenues at 14 billion a year and Claude Code alone is pulling in 2. 5 billion Elon Musk called Anthropic misanthropic and evil worth noting Musk runs a competing AI company and Anthropic blocked X AI from using...actually putting Claude to work. For 6 months, Goldman has had anthropic engineers inside their offices, building AI that checks trades and onboards new clients. They brought Claude in for coding. Turns out it could handle compliance, too. When the biggest investment bank on Wall Street embeds your team for 6 months, they