over to Visual Studio code this is super powerful for building stuff outr if you want to get access to visual studio code go to code. visual studio. com then navigate to visual studio inside the extension section over here you want to download Cent so make sure you have Cent downloaded which you can see right here and then...want to make sure you have that selected as your API when you're building this out now you can see that the code is being done on the right hand side and Claude 3. 7 Sonic is at the point where it can almost code any SAS tool that you want right within reason as long
Поиск по транскрипциям
benchmarks. Claude 4, Claude Opus 4, most powerful model yet, and the best coding model in the world. That's a bold statement, but you can see how it's performing and why they've justified that. What else we got here? So, cursor calls it state-of-the-art for coding and a leap forward in complex codebase. Replet...coding right if we look at these and compare them that was at 63. 2% for software engineering now we can see Opus 4 is scoring 72. 5 and Sonic 4 is scoring 72. 7 compare the benchmarks over here you can see them and how they perform. Wow look at this as well. So this is Claude playing Pokemon
build stuff with it this model can understand what you're asking fix mistakes and make sense of complex data plus it has top-notch coding vision and writing skills the benchmarks are crazy Claude 3. 7 Sonet is beating Claude 3. 5 Sonet open AI deep seek and grock 3 beta in multiple categories on MML U is scored
Today, Open AI is the one hitting the panic button. — Sam Alman sent an internal memo with two words in the subject line. Code red. — Code red. — Code red. — That's their emergency signal. — Get out. Run. — It means drop everything. We have a crisis. And here's what caused it. Google just released an AI that's beating Chat...match that distribution. Front two, Anthropic, the company behind Claude. There's a test called SWE Bench that checks if AI can fix real coding bugs. The same bugs human engineers deal with. Claude just scored 80. 9%. First AI ever to break 80%. OpenAI is best around 77%. That
focus. Version 4 is being positioned as a coding first model and internal testing by Deepseek employees suggests it could outperform rivals like Entropics Claude and Open AI's GPT models specifically on coding tasks. One of the biggest reported breakthrough is versions 4 ability to handle extremely long coding prompts which could be a real advantage for developers working...global stage. If these claims hold up, Deep Seek version 4 could be one of the most serious coding focused model launches we've seen this year. Just to show you guys where Deep Seek currently stands versus Claude Opus 4. 5. I've asked both these model to make a Voxil style world, add trees and a village
Make the screen crazier with more animations, music, flashing colors, and overall craziness. plus made the screen shake. There we go. All right, because Claude 4 is so fast, like we can probably code two versions before the other two are ready. So, Gemini is still coding back here in the code back from AI Studio. You might be wondering
have a bunch of code here, including our specs and to-do. And there's a bunch of code in this folder. So now our code is in GitHub. And we can uh ask Claude to, you know, revert to previous version or, you know, add new commits as we move along. All right. So now, what are we gonna...about this tip, which is to clear context strategically with clear and compact. Because we spend all this time going back and forth through Claude in a chat window, Claude's contest window is probably pretty full by this time. And AI just gets progressively worse as your conversation history gets longer. So let's actually type in compact here
model you can run locally as you can see and if you don't want to use claw code or you don't use claw code then just copy and paste the GitHub into claude and you can go from there
Make the screen crazier with more animations, music, flashing colors, and overall craziness. plus made the screen shake. There we go. All right, because Claude 4 is so fast, like we can probably code two versions before the other two are ready. So, Gemini is still coding back in. And then we've got the code back from AI Studio
here's the game changer, desktop extensions. These let you connect Claude to your local tools and files through something called the model context protocol. Claude can now read your project files, run your code, and access your databases with your permission. Of course, the setup is dead simple. Download from claude. ai, AI. Install it, sign in with your
versus Claude 4 versus Gemini 2. 5 Pro. I just ran China's new version of Deepseek R1 against Claude 4 and Gemini 2. 5 Pro and the results have shocked me. This is a brutal AI coding battle and what happened next is going to be very interesting. I'm about to show you which AI destroys the competition...limit and see which one provides the most professionalgrade code that actually works. And you'll discover which AI model wins the most intense AI coding gauntlet I've ever created. Deepseek R1052A versus Claude 4 versus Gemini
life in search. This is available to everyone in the US in English right now. But basically, it appears to be the same kind of features you get in Claude or OpenAI where if you ask it to build something for you, you can actually see what it built inside of the canvas. So, if I go to Google...game in HTML and JavaScript code. And we'll see over on the canvas side, just like you would expect inside of Anthropics Claude or ChatgPT, we can see it, write the code, and then we can actually demo the game right here inside of Google's AI mode
Well, it only works for tech people. " But the truth is like I can build a tool and anyone now without ever having coded anything, right? You just describe what you want. Claude built it and you can use it right now for creating blog posts, for writing landing pages, for coding tools like you can see we're doing...easy. So how do you get Claude Genspark to secure the web page? That's a great question. I'm not a security expert to be honest with you. But the way that I would approach that is I would say okay you know now advise me plus update the code on this page be more secure right something like
code with each of these if you want to use the API whilst waiting for these to load? So Gemini is the cheapest to code with. You can actually get a free API. Claude, Sonnet. Let's compare these two. So we can actually do a comparison on Open Router in terms of which one performs the best. We open...Gemini, doesn't work. But 03 actually performs better and it's cheaper on the API versus 03 versus Claude Opus, right? So big difference there. And you can see actually 03 is not a bad model to code with at all. So let's try something else. Now what
Claude can't automate its own self-improvement just yet, but what about on a more practical level? Now that they're releasing, for example, Claude in PowerPoint and now that Opus within Clawude Code is writing a significant fraction of the world's code. Let's start with generalized measures of knowledge work, GDP val. And frustratingly, Anthropic
notice cloud code make a mistake, then I tell it to find the relevant Claude MD file and make a change there instead of the root one. I try to have my Claude MD files as short as possible. And every time there's a brand new model update, then I look at what I can remove. I very rarely...hype threads where they just talk about like, oh, you should add all of this to your Claude MD file. I do share more Claude MD best practices in a free video in my mastercloud code class if you interested. Next up is I noticed that
Today I'm going to show you how to turn Claude into a worker that never stops. It's called Ralph Claude Loop and it's nuts. This thing keeps coding until the job is actually done. No babysitting, no handholding, just results. It's brand new and developers are losing their minds over it. This is going to change
specialized tasks, Claude. md files tell Claude about the specific project. Things like your text stack, coding conventions, and repo structure. CloudMD files live alongside your code in the repository. A CloudMD file may say things like we use Nex. js JS and Tailwind. But skills on the other hand are portable expertise that work across any project
back to a new chapter here, we've got 2. 5 Pro experimental and then we've got Claude, which I would say is the second best option, right? So, you got two different options there. Again, this is for coding specifically like this is what Gemini is really good at. So, if we go into for example here...much detail as we want inside this project and then deploy it as well in a subdomain. but it's miles above the second best coding tool in the world. And I think until Claude releases like version 4, it's just not going to be able to compete. It's nowhere near the same level. Now, the other cool
back to a new chapter here, we've got 2. 5 Pro experimental and then we've got Claude, which I would say is the second best option, right? So, you got two different options there. Again, this is for coding specifically like this is what Gemini is really good at. So, if we go into, for example, here...much detail as we want inside this project and then deploy it as well in a subdomain. but it's miles above the second best coding tool in the world. And I think until Claude releases like version 4, it's just not going to be able to compete. It's nowhere near the same level. Now, the other cool