human teammates can focus on higher leverage stuff, the creative thinking, the product strategy, the work that actually moves the needle. So, OpenClaw or Claude with Claude code and claude co-work. Now, we know that both ecosystems are evolving fast. So, what I want to share here is just a snapshot of where things stand right now here...insights from the trenches. You can get yours by going to buildermethods. com. And members of builder methods pro get access to my full course on building with cloud code and my new series I'm rolling out on how to build an agent team with OpenClaw. You can join us and get access to both by going to buildermethods
Поиск по транскрипциям
performs the task, but you don't necessarily inspect the internal logic. Just like how Gemini 3. 1 may have found spurious patterns in ARC-AGI, in your code base, Claude or Codex may overfit to the spec, or may drift from your original concept. So, the fallibilities presented in this video are relevant to you even if you only...care about coding, or letting your open Claude agents code for you. Gemini 3. 1 Pro indeed hit a record Elo in live Code Bench Pro, which involves competitive coding problems. That's great, but you can turn that optimization dial a little too far. Let me show you what happened last night when I used Gemini
check if you got what you wanted. And I mentioned this also in the Claude Code video, and it's very important to me, you're still allowed to write code. You're still a developer. Of course it's shifting. Of course it's more about planning and controlling and steering the AI and doing code reviews. Our role...code, at least I do. But you're still allowed to write some code. And then there is one last thing I also wanna mention, which also applies to Claude Code and essentially all these coding agents, but also to Codex. You have this concept of a Agents. md file, right? And the idea here really is that you have
rang the bell on the NASDAQ, right? And so I'm super proud of that. Like, I actually you know, the other thing that's interesting going back to Claude Code, like when if people apply to YC now, we actually now give people the option to upload a transcript of that their chat with Claude as they vibe coded...that way, do this way. " But like, I think that that's actually a very interesting equalizer. Like, your ability to actua-- like, it's, you know, the, the Claude code chat we think is a little bit like the SAT. It's, it just, like, you can look at what, how someone prompts and you can understand, like
using Claude Code right now. In a second, I'll show you the best way to be using it, but I want to go over all the ways you can be using it right now. You can be using in the CLI. So, you just open up your terminal. You could be using the CLI and cursor, the extension...cursor. I actually think these two ways are the most popular ways people are using Claude Code. And it's actually the wrong way to be using it for many different reasons, and I'm about to explain to you why. Yeah, the mobile app, the desktop app, and then you can be using Claude Opus in the hundred different
show you how this actually works. It's stupid simple. You open Claude Code on your desktop. You click schedule in the sidebar. You hit new task. Then you fill in four things. a name, a description, the prompt you want it to run, and how often run. That's it. You click run now to test it, and boom...time, Claude launches a brand new session. It runs your prompt and then it sends you a notification with the results. Every single run is a full Claude code session. That means Claude can edit files, run commands, create commits, open pull requests, even use MCP tools and APIs. This is not some watered down version. This is the full
going to show you how to set it up, how to use it, and we'll see just how powerful it is. Does it match up to Claude Code? We're about to find out. If you stick here, you're going to be a master of Gemini CLI. I promise you that. So, what I'm going...most out of this. So there is no planning mode with Gemini CLI. That's a downside here. Cloud code has that. In Claude Code, you can just hit shift tab and you go into plan mode where you just go back and forth with the model to build out a plan for your app. That doesn't exist here
Claude GitHub app, and now Claude has access to all your repositories and can write code for you on the go. The way this basically works is Claude is not writing code on your computer. It's writing code on the cloud and then storing it in your GitHub. Then you have the ability when you get back to your...listed in the cloud code for web. So you can start doing work on all of them if you want. So, let's talk about 10xing your productivity with Claude code
clone said insane? That's a real question. It said it a lot, my friend. All right. So, you want to see something here? I actually thought that the claude code wouldn't be able to do this because I know I'm 99% sure like other stuff like chat GBT or Gemini would fail at this task so much...Claude would absolutely crush it because we got the new update from Summit and that powers Claude 2. 0, right? Claude Code 2. 0. So, if we open this up, this is the MR game and we went from idea to game in the space of what, five minutes. It's insane. Like it that is genuinely insane. John
After spending over 800 hours in Claude Code, I've discovered ways to actually make this thing work for me, not against me. Especially at times when I need to move fast as a solo developer. Whether you're churning through tickets at a corporate tech job or attempting to vibe code your next 10K a month app. If Claude...setup. In this video, I'll walk you through my private workflows, essential resources, and teach you the six core features that most developers are unaware of in Claude Code. So, by the end, you'll be shipping faster with Claude instead of wasting time fighting it. Before we dive into
everything to do with Claude Code 2. 0. Anyways, moving on to Sonnet 4. 5. You can see right over here, it's now the best code like software engineering agent in the world. So, it gets a high score on the SWE-bench Verified benchmark over here. You can see some of the scores over here, and it seems...more interested in all these scores, which is like a super long document which I won't be reading. As I said earlier, they renamed the Claude Code SDK to be the Claude Agent SDK. So, you can now basically build agents for any other task quite easily with this SDK. And they also released this new research preview
videos, I recommend using GitHub as much as you can. You can say make me a slash command that commits and pushes all our code to GitHub. And what Claude code will do here when I hit enter is actually create a custom slash command for me that will push all our code to GitHub when we are done coding...code on to GitHub. Make custom slash commands. It will save you so much time. And you don't even have to write them yourself. Just tell Claude Code to write them for you. And if you can't think of any slash commands that save you time, just go in here and say, "Hey, what are some slashcomands that
every single movie without any limit. That's a memory killer and API spammer. You can't ship something like that to real users. When I compared it to Claude code, Claude won on usability. Claude's UI might not be as pretty, but it works perfectly on the first try. That matters more than looks. Test number...multimodal features are genuinely useful. Being able to show its screenshots and get working code is powerful, but the reliability issues hold it back. When it works, it's impressive. When it doesn't, it fails hard. I'd still recommend other models like Claude for serious work. They're more stable and consistent. But Supernova is worth keeping
with 1 million context window. And it's supposed to be coming very soon. And Claude Code is also supposed to get an update where your agents, the various agents you might have, will be able to talk to each other, which is kind of crazy as well. And then he's really confident that this model is going...models. So, we got multiple people talking about a release that's coming very soon and claims about superior performance, mysterious code names, and updates to Claude code. So, there's a lot to unpack here. But wait, there
Okay? So, you're going to make a good account. free Versel account, both free. You're going to give the links to Claude Code in your terminal and say, "Hey, connect to these. " And then every time you make a change or an update, you're just going to say commit. And then it exists on the web like...lives on the internet. Congratulations. It's that easy. Okay. So, regular claude, I told you claude code, just listen to me. Just use it. Now, here's how you install it because this is a bit confusing if you're not technical at all. So, we're going to go through Mac. Um, number one, you're going
claude code within Visual Studio Code. Um so you can see here I'm just going to it looks very similar to anti-gravity. I'm just going to open up our executive assistant folder here. And the way that I see the future of this all going, Greg, is I think that everyone's going to have their what...Google Drive, Calendar, Granola for my meeting notes, Stripe for payments, notion for project management, and I don't even enter these tools anymore. I just sit in Claude Code as one central place. And an example here is I sent myself before the episode, I sent myself an email from a fake prospect and I also entered in Granola
open terminal. If you're on Microsoft, you open up the command line and then you just type into the terminal what you want built and then claude code goes off and builds it for you. What I'm going to show you in this video is actually how you can open it up inside a cursor so that...goes and it just does a tremendous amount of work. Well, if I was using cursor by itself, the agent built in, I would sit there and babysit. Claude code can do a lot of things independently. It's also built on top of the most powerful coding model that exists, which is Claude for Opus, which is 10 times
enter and we'll see which one performs the best side by side. Now, I have a feeling the Gemini 2. 5 Pro is actually going to code a lot faster than Claude, but we'll see which one creates the best output side by side. All right, so Claude is on the left hand side here using Opus...including like some of the best automations with AI3 and all sorts of stuff like this. So let's see how they perform. Now we've got Gemini coding out. Claud'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
give 03 the head start and we'll test this out. Now the first thing that I'm going to say even before this is finished coding out right is that inside claude you actually get artifacts which means that you can embed and see the canvas directly inside here whereas with 03 you don't get the canvas option...code back from chat GBT now. So, Space Invaders on Chat GBT dopamine Space Invaders has coded out way faster there than it has directly on Claude. But we can run the code here. So, let's run the code. I'm just going to mute this tab in case it starts blasting out crazy noises. What just happened? What
Google studio, you get it on GitHub, then you can start editing with cloud code in like 10 seconds flat. So that is really convenient for editing with claude code if you want to take it to cursor, claude code, codeex, whatever you're using. Tons of the cookbooks available so you never have to come up with ideas ever