Spend 1 minute every day to get the latest AI information.
The content includes but is not limited to cutting-edge AI news, AI tools, AI painting, open-source projects, and learning tutorials.
The main feature of the brief is its concise description, but important information is still introduced in detail through independent posts.
Here is the latest AI information for June 25.
Cutting-edge News
1. Many developers received warning emails from OpenAI!
Starting from July 9, OpenAI will strengthen the blockade on the use of its API in unsupported regions.
Personal speculation: Those who received the warning email probably accessed the API through domestic networks by mistake, which was recorded by OpenAI.
If you use the same account for API access and daily use, it is recommended to back up your data, such as created GPTs, commonly used GPTs, and related prompts.
2. OpenAI acquires Multi.
Multi is a multi-person real-time collaboration product, mainly used in multi-person interactive video calls.
Will OpenAI integrate AI into real-time video to achieve more functions through this acquisition?
Announcement: https://multi.app/blog/multi-is-joining-openai
3. GPT-4o real-time voice feature finally starts limited release!
A user on platform X received an invitation to use GPT-4o Voice when opening the iOS client.
However, according to the author, it is currently just an empty UI interface and still cannot connect to the model, not much different from the previous voice function.
The author also wrote a detailed article about it, which you can read if interested.
Article: https://www.testingcatalog.com/here-is-what-to-expect-from-the-limited-alpha-for-chatgpt-voice-mode/
4. Sony, Universal Music, and Warner jointly sue Suno and Udio.
They accused the two AI music generation companies of copyright infringement and demanded $150,000 in compensation per song.
News source: https://www.vulture.com/article/major-labels-music-ai-suno-udio-lawsuit.html
Open-source Projects
1. An agile information mining tool Wiseflow.
It helps you crawl relevant information from websites, public accounts, social platforms, etc. You can filter, refine, tag, and upload the crawled information to a database automatically.
You can even easily integrate it into any Agent project as a dynamic knowledge base.
GitHub: https://github.com/TeamWiseFlow/wiseflow
Moreover, the tool can be fully deployed locally, requires minimal hardware, no GPU needed, and is suitable for any hardware environment.
2. Streamer-Sales, a large model for sales anchors, aims to make you a sales champion.
It is a sales anchor large model that can explain products based on their characteristics, generating fine and unique commentary to stimulate users' purchasing desire.
GitHub: https://github.com/PeterH0323/Streamer-Sales
Feature Summary:
📜 One-click generation of anchor scripts 🚀 KV cache + Turbomind inference acceleration 📚 RAG retrieval enhanced generation 🎙️ ASR speech-to-text input 🔊 TTS text-to-speech output 🦸 Digital human commentary video generation 🌐 Agent uses network to query real-time logistics information
I hope this sales anchor large model can become an indispensable helper, enhancing your sales and user experience.
The project is fully open-source, but deployment is somewhat difficult and requires high configuration. Interested parties can try it out first.
Online experience: https://openxlab.org.cn/apps/detail/HinGwenWong/Streamer-Sales
