minute, and bakes it right in without all the fluff, keeping the agent side of things at the forefront. And I think what happened is when Claude Code brought the AI in the terminal to the IDE, it completely eliminated that context switching that we all hate. And that's what warp is recognized as well. No need to leave...building something and then you make all the little tweaks you need by hand. So, that's a huge improvement and I think brings it to par with where Claude code is, but with its own particular flavor and perspective. You don't integrate one with the other. Warp actually just has it all in one package, one single environment
Поиск по транскрипциям
long as you like. Now, let me quickly show you the pricing. So, the $20 monthly plan, if you pay monthly, that gets you Claude Co-work and Claude Code, too. But again, it's going to be limited on how much usage you get out of it. I actually quickly hit my limit here. This is the plan...dependent on usage. Starts at $100, but right now I believe I'm paying $200 for that. And they also have T-Mn enterprise plan that also unlocks Claude code and claude co-work. Now, hopefully it's really obvious that Claude Co-work is one of the closest updates in AI that I've ever tested that feels like
this using the subagent spec. If you don't know what that is, do a little bit of research on subagents. It's an anthropic and Claude Code feature that's quite well supported by our current workspace structure. And once you've built the subagent using a sonnet-4. 5, I want you to roll...gonna do is actually spin up a bunch of subagents to do research. So that's what this task little bubble is, right? When it says research Claude Code subagents, what it's actually doing is it's giving this task to a subagent called the research subagent. There's another subagent as well, like the search subagent
really interesting watching it go. Okay, so it seems that both of them are done. Codex CLI did use more tokens as well, it used 130,000 tokens, whereas Claude Code used 72,000 tokens. So I expected more tokens, it probably got more context from online or something, so I'm hoping the solution will be better...with a slightly different like, padding, color scheme, and design. But yeah, as for the underlying logic, it works really well. And now this is the one that Claude Code gave us, so we can go to MCP servers. It did copy the design over, like the design is very similar to the one in models over here
actually an amazing vibe coder. It can control any vibe coding service you use. So the best ones to use are through the CLI. So if you have Claude code or if you have Codeex, make sure to get that installed on your new computer and have your Claudebot use it. So you can say, "Hey, I'm using codec...code and then give it ideas for things to build. So I told it to build this canband board. It wrote all the code for me through claude code. It uploaded it to GitHub. It created a pull request. It sent me a link for the pull request and I was able to review it when I got home. This
just open this up and see if it works. All right. So this is the version from open work as you can see. And this is the version from Claude Code, right? From Claude Co-work. Honestly, if I had to choose one, this is a tough one, but I'm still going to say that Claude Co-work...right here, it still did a similar result for the Tetris game and it can still do similar things. So it may not be as good as Claude Code and Co-work, but it's not far off, right? But you can see the speed of it. So, for example, if we have a look side by side
transcript from this podcast if it was afterwards and I would give it even more context, turn that into a lead magnet. Of course, I would then move to Claude Code to create the lead magnet landing page and we're done. And so for me, this process used to be like me chatting with my team, which I still...start with the cloud interface like the Mac app — is interesting. — Yeah. I Okay, — I'm getting better at moving or knowing when I want to move between Claude code and uh Claude. One thing I will say is so a friend of mine who's running a company called Pipex, they're like he's gone thousands of levels
look like when you open it up. Now, what you have to do is come over to the lefth hand side, click on extensions, and you'll type in Claude Code, and it will look like this. And all you have to do is go ahead and install this. When you do that, it'll prompt you to sign...with your Enthropic subscription, which you do need to be on a paid plan for Claude Code. You could use your API key, but it's better to just have a fixed cost rather than worrying about how many tokens you are racking up. All right, so once Cloud Code has been installed, you'll notice in the top right
this is V1. Here we go. Uh, from a looks perspective, I mean, it's comparable. I think I lean a little bit more towards the Cloud Code one. From a looks perspective, let's see from a data perspective, that's accurate. 201108. It oneshot the functionality. It oneshot the AI fun. So 2011, boom, that's the actual...sentiment. But the price perspective, it had in one shot, which is pretty amazing. From a look and feel perspective, I think it's quite a bit weaker than Claude Code. Claude Code
started. Claude Code can now control Claude in Chrome. So if you don't know already, this is a browser extension for Google Chrome and Chromium based browsers. So when you sign into your account and click on the button, then you can see you have claude on the right hand side here and you can do like any basic...tasks like getting it to summarize a page and so forth. But the interesting thing here is that now claude code can connect and control your browser via this extension. So firstly you want to make sure you're on the latest version of cloud code by doing cloud update and then after that you can run cloud code
sandboxing in Claude Code and have your remote control sessions run through that sandbox. So I cover much more about sandboxing in my master Claude Code class that will be linked down below if you're interested. But to give you a brief overview, if you make a folder like this and inside of your settings. local JSON file...research on. And then I can deny any tools such as editing or writing tools, for example, and bash tools. And if I start a remote session of Claude Code here by doing claude rc, I can go to a session which is in the sandbox research environment here and then say, go on archive and find me the latest
actually build. So let's jump into this first phase of the build and public framework which is the actual building part. You need to build something with claude code. So the first step in the process is building an app. The best apps in the entire world solve problems. So if you want to build an app people...basis. Every challenge you have from beginning today to end today, write it down. Then, what we can do is we can go into Claude Code and find solutions to those challenges. So, now that we have Claude Code installed, I'm going to type in Claude in a brand new folder on my computer
other files, but it figured out, okay, which ones are screenshots, and then it deleted them. So, it is ridiculously fast. I actually heard a crazy story that claude code coded out the whole of code, all of it, in a week and a half by the claw team, which is insane when you think about it, right? So, from...what this can actually do is use your browser. Right now, the cool thing about this is it's really targeted at people who are not technical. I think Claude Code put a lot of people off because it's a coding tool. Whereas this is like more non-technical, more chilled to use, more easy and very simplified, right
that AI will be involved in almost all of the code that gets written. But still your role as a lawyer or as a developer in understanding what the code needs to do and having taste and guiding what is done with the software is going to be more important than ever. - I mean already at Cursor it's probably...like a PR and he's like asking for a stamp and you're like, what are you doing? And it's like, yeah, he's using, you know, Claude code or some coding tool with Claude has the base model there to like fix bugs in a production code base. I think that's amazing as well
what he can improve. Thank you for the $10 Canadian. Slender Morris Furnishings with the 799A. Australian. He's Australian. Slender Morris Furnish is Australian. Do you use Claude code, not Clawbot to build mission control? Um, no. Claudebot built all of mission control. All this is by Claudebot. Henry made all this. I still use Claude code for bigger...boot camps every Friday. And you get instant access to all the previous live boot camps uh are all recorded and put in the vibe coding academy plus full claude code master class plus full uh social media X master class. Which Mac studio did you get? M4 Max or M3 Ultra? M3 Ultra. Two M3 Ultras. And then Exolab
next tip, which is a critical rule you need to put into Claude Code if you are a new user of Claude Code or new to Vibe Coding. And that is this rule right here. After completing a task that involves tool use, provide a quick summary of the work you've done. This is very important...building features, writing code, and all that. And a lot of the time, it's so eager that it's not going to explain to you actually what it did after it uses different tools and does complex things. So, if you're a new vibe coder and you want to understand everything Claude is doing, which I highly recommend
coding one, which is crazy. So obviously Anthropic has figured out something related to coding the others haven't. And in practice, developers universally agree that Claude writes functional code on the first try more consistently than alternatives. Here's a real world example. I needed to bulk export conversations from a customer service platform, but their support team said...Claw to generate mermaid code, which you can then paste directly into tools like Excaliraw to get clean visuals in minutes. Area two, polishing copy. Beyond code, Claude produces written drafts that sound human and need fewer revisions. When you need to tighten an argument or match a specific voice, Claude just gets it. Put simply, it's exceptionally good
actually not really a planning mode. Planning mode is just a take a long time mode. And what it actually does is it writes code. Planning mode and claude code or cursor, it just creates a plan and then you have to approve it. But the thing with planning mode in anti-gravity is it actually does the coding...really impressed with the speed of Google anti-gravity here. One issue and challenge I'm having with codeex lately is it's a little slow for my taste. Claude code always been kind of slow, but it seems like it's been worth the slowness with the really nice results you get. Composer one in cursor extremely fast
Most people are visual learners. I am myself. You probably are watching this video. The problem is coding agents like Claude Code are not very visual tools. When you want it to explain something to you visually or even just help you explain something to someone else, they really don't do a good job unless...build dozens of these every single month. And so I've taken my entire workflow and I've packaged it up into a skill so I can use Claude code or any coding agent to help me build all these beautiful and practical diagrams. And so giving my coding agent the ability to argue visually. And so right
like a tree. You have different branches. Your main branch is obviously your main branch and you're taking this code from all the other sandboxes and environments that claude code for web and it is smooshing it into the main which is basically your main working branch. So that's how you pull down the code, review...road maps tweets to review, and my app literally got better while I was sleeping. That is how you take your productivity to the next level. That's why claude code for web is so unbelievably powerful. You now have the ability to spin up an army of AI agents. You now have the ability to be a oneperson business