Introducing "Make it Heavy" a framework to simulate Grok Heavy functionality, fully open-source and running in your terminal.

Use any model you want, thanks to @OpenRouterAI. 👇 Link here:
github.com/Doriandarko/ma…

Introducing MagicPath, an infinite canvas to create, refine, and explore with AI.

Create beautiful components and functional apps, while providing production ready code.

Available today, free, for everyone.

The Cursor moment for design is here. In @Magicpathai every component is a conversation.

Edit and refine, either manually or with AI.
Design is the art of keeping good decisions. MagicPath lets you copy and paste interactive components across the canvas to explore and perfect your ideas.

Introducing RAT:

Retrieval Argument Thinking.

Extract just the reasoning process from deepSeek-r1 and send it to any LLM via @OpenRouterAI.

Boost the second LLM's performance and gain access to missing capabilities like function calling and JSON mode. 👇 🐀🐀🐀
github.com/Doriandarko/RA…

If you're a CS student worried about recent AI developments, read this. First, intelligence is not a measure of effectiveness. I've managed many people throughout my career, and if you don't know how to employ their intelligence effectively, raw ability simply doesn't matter.

Introducing Claude Engineer V3 🚀

The first self-improving assistant capable of not just running tools, but creating them on the fly whenever needed!

Want a web scraping tool? Just ask Claude, and it will build one for you. This is my most ambitious release yet. I started exploring self-improving frameworks a few months ago, and after lots of usage and testing, it does feels like the future of computing.
github.com/Doriandarko/cl…

Today @Anthropic is releasing MCP, a framework that allows Claude to run servers, giving it superpowers and effectively turning the Claude app into an API.

We created some server that I think you'll love!

FileSystem: Claude can create, read, and edit files and folders locally. Brave search: Claude can search the web for anything, and provides enrichments for location search results. leveraging Brave web search and local search api.

o1-engineer is here! 🚀
A coding assistant built from the ground up to leverage o1 reasoning capabilities.

It can create and edit multiple files or entire folders,
plan complex projects, execute them, and write code reviews.

All from your terminal 👨‍💻 💬 You can chat regularly or execute commands:
/create to create files or folders
/edit to edit file, files, or folder content
/add to add files or folders to the chat context
/planning to create detailed plans
/review to create code reviews that you can use directly

Here are some o1 prompt techniques I’ve discovered that have worked well for me so far and can likely be applied to any reasoning LLMs.

1. Be explicit about the reasoning chain
By requesting a detailed breakdown, you encourage the model to organize thoughts methodically.
2. Guide decision-making with hypothetical scenarios
Providing specific criteria guides the model through a structured analysis, yielding a more nuanced response.

I'm excited to share that I've been partnering with the @AnthropicAI team to build a beautiful, fully customizable customer support AI agent with RAG.

Using Claude, NextJS, ShadcnUI, and AWS Knowledge Bases.

The entire repo is available on GitHub via the Anthropic Cookbook. ✨

https://twitter.com/alexalbert__/status/1831022348784542090

It's a great starter app for anyone to really take advantage of Claude's capabilities.

You can see Claude thinking in real-time, as well as the mood of the user interacting, if any documentation has been used, and even the category to which the customer problem is connected.

So many of you have asked, and now it's here 🗣️

Introducing Ollama-Engineer. 🦙

Experience the power of my original agentic coding assistant locally and completely free via @ollama on supported models.

Llama 3.1, Mistral Nemo, and more. Repo
github.com/Doriandarko/cl…

Introducing Claude Engineer 2.0, with agents! 🚀

Biggest update yet with the addition of a code editor and code execution agents, and dynamic editing.

When editing files (especially large ones), Engineer will direct a coding agent, and the agent will provide changes in batches.

Batches are smartly selected based on file complexity.

The code execution agent will run the code and check for issues.

It can even start processes (like live servers) and end them.

It's insanely powerful! 🔥 You can now save your chat by typing "save chat," and it will create an markdown file of the entire conversation.

This is perfect if you are looking to resume a conversation another time just by providing this context.

You can also reset chats.

Claude Opus's ability to orchestrate subagents is absolutely insane and deserves more attention.

Watch Claude direct subagents to build an entire drawing app. 🧙‍♂️

If you like, I could share this code, which can basically solve any goals you present, step by step.

Like Devin! This is the full output saved as a text file, it's fascinating to see how Claude thinks.

The whole process took exactly 13 minutes and 20 seconds.

pastebin.com/F4tMW8nQ

Claude 3's retrieval ability over long content is so good that if you provide structured data, it essentially acts like a fine-tune! 🤯

I gave it 800 of my tweets and now it just speaks like me. It's hard to convey my level of shock...

It even tagged @Replit and @everartai! For context, I have fine-tuned hundreds of large language models; these are the best results I have seen on my tweets.

AI influencers are set to be a major trend in 2024. ⚡️

@everart is the best place to create a consistent AI character.

Just a few photos – real or AI-generated – and you can create a unique AI influencer.

Let's see how to do it!
Meet Emily! 👇 First, head over to the "New Style" page.

Start by selecting "Person". This step enables @everart to work its magic in the background.

Ever wondered how to fine-tune SDXL for any style you desire?

I've got the simplest guide for you using @replicatehq.

Let's dive in! Why fine-tune?
Fine-tuning lets you specialize a model to specific styles, people, or objects, making it perfect for generating high-quality, consistent images. SDXL's capabilities are game-changing.

Here an example from one of my model trained on Hiroshi Nagai art.

Many people have asked me for a tutorial on how I built Zing (Bing AI clone).

Follow me in this step-by-step and how with a bit of nudging and clever prompting, I got GPT-4 to do that for me in around 12 messages.
🧵👇 First GPT-4's knowledge still stops at 2021, which means it has no idea of how to implement @OpenAI's new Chat API. So, I just provided it with the entire API documentation in the first prompt.

I built a Bing AI clone called Zing!

Featuring functional GPT-4 and Bing API search integration.

I didn't write a single line of code;
I just asked, and GPT-4 did the rest 🤯😅

It took roughly 2 hours.

Hosted on @Replit Unlike Bing, Zing still has a sense of humor

I am 99% convinced they are using the #DALLE2exp version here. I ran some comparisons and the style matches pretty well. Check out the pics in the thread - the Exp is on the left and Bing is on the right.

https://twitter.com/jordirib1/status/1638167237856235520

Why spend time making GPT-4 prompts when it can make the best ones for you?

Introducing God mode.

Here's how I did it 🧵👇

#GPT4 #AI #GodMode First I asked GPT to be the best prompt generator on the planet.

I used #DALLE2exp and Bing AI to create Instagram posts for a fictional estate in Costa Rica.

In order to achieve a consistent look, I instructed Bing first and then added some personal touches.

You have to see them to believe them.📸🌴