Spend just 1 minute daily to get comprehensive updates on AI technology developments, industry trends, and market insights.

The content includes but is not limited to cutting-edge AI news, AI tools, AI painting, open-source projects, and learning tutorials.

Stay tuned to AI Daily for the latest in AI trends. For important information, detailed posts will be made separately.

Here is the latest AI information for July 9.

Cutting-edge News

1. Poe launches a new feature Previews, similar to Claude's Artifacts.

Allows users to generate web applications directly in chat and interact with them in real-time, such as interactive mini-games, data charts, and more.

Conversation example: https://poe.com/s/neKbyqEJYUtiRjeC6b5f

Claude and Poe have successively launched real-time code preview and interactive functions, which will attract some ChatGPT users. OpenAI needs to catch up!

AI Tools

1. The most comprehensive price comparison tool for different LLMs: llm-pricing.

On this platform, you can choose various model versions and fill in the input and output tokens to calculate the API cost in real-time.

A handy tool that can help you choose the model that fits your budget.

Online usage: https://huggingface.co/spaces/philschmid/llm-pricing

AI Painting

1. An ultra-high-resolution image generation model: UltraPixel.

Supports generating ultra-high-resolution images (1k to 6k) directly while maintaining image generation efficiency.

Detailed introduction: https://jingjingrenabc.github.io/

From the demo images, it retains a lot of details, such as hair strands and the roots of plants.

The model is not yet open-source, so keep an eye out for updates.

2. New project by ControlNet author, Minshen: Paints-Undo, fun and interesting!

Can generate the drawing steps of an input image, simulating the human drawing process from sketch to completion.

Includes sketching, outlining, coloring, shading, adjusting color curves, changing layer visibility, and even altering the overall composition during drawing.

GitHub: https://github.com/lllyasviel/Paints-UNDO

Interested users can try it on Colab, though it requires Pro for smooth operation.

Experience link: https://colab.research.google.com/drive/1_WCZUXUP3xeZar8v63E26wDKlMBUAPO-?usp=sharing

Open-source Projects

1. A Python library for creating high-quality tables: Great Tables.

It has a rich set of table components, such as headers, footers, row labels, column labels, etc., and provides various formatting options to help you create various types of tables.

GitHub: https://github.com/posit-dev/great-tables

With this library, anyone can create beautiful tables using Python.

2. An open-source GitHub Pages template: academic-homepage.

Helps you quickly set up a personal academic homepage, including basic content such as personal introduction, education experience, and published academic lists. The page layout is also mobile-friendly.

GitHub: https://github.com/luost26/academic-homepage

Learning Tutorials

1. An open-source deep learning guide on GitHub: "From Tensors to Stable Diffusion".

Provides a 9-week deep learning course syllabus, from the basics of tensors to the implementation of stable diffusion models.

Divided into 5 parts: Introduction to Tensors, Basics of Deep Learning, Visual Models, Language Models, and Implementation of Visual Language Models.

GitHub: github.com/jla524/fromthetensor

The course also includes video explanations, code implementations, and related paper studies.