Today's featured AI news highlights, enjoy reading 👇

🎓 Google launches new education model LearnLM 1.5 Pro Experimental, now available for free trial on AI Studio, dedicated to providing personalized learning experiences.

🗣️ ChatGPT opens advanced voice features on web for paid users, supporting multiple languages including Cantonese, enriching voice interaction experience.

✈️ AI Travel Agent, an intelligent travel assistant developed based on LangGraph, integrates multiple language models to handle travel planning tasks like flight queries and hotel bookings.

💻 ScreenPipe open-source AI desktop application supports offline use, can monitor computer activities 24/7 and provide intelligent dialogue and summaries through LLMs.

Latest News

1. Gemini launches new model LearnLM 1.5 Pro Experimental.

A new model fine-tuned for learning and educational research, aimed at making teaching and learning experiences more engaging, personalized, and attractive.

Try it online: https://aistudio.google.com/

Currently available for free trial on AI Studio, interested users can go experience it.

2. ChatGPT's advanced voice mode is now available on Web.

OpenAI is rolling out web access to advanced voice mode for all paid users, go check if you're covered.

Official website: https://chatgpt.com/

By the way, while playing with it, I discovered it can speak Cantonese quite well. Interested users can give it a try.

Advanced Technology

1. An AI Travel Agent built on LangGraph: AI Travel Agent.

Handles travel planning tasks using multiple language models for finding flights, booking hotels, and sending personalized emails.

GitHub: https://github.com/nirbar1985/ai-travel-agent

To run this project, you need to configure keys for OpenAI, SERPAPI, LangChain, and SendGrid, and it provides a chat dialogue interface.

In the chat box, send "I want to travel from xxx to xxx, October 1st to 7th, help me find flights and four-star hotels" to get relevant content.

Open Source Projects

1. A powerful offline open-source AI desktop application: ScreenPipe.

It can monitor your computer 24/7, collecting information through screen recording, OCR, audio input, and transcription, saving to a local database.

GitHub: https://github.com/mediar-ai/screenpipe

Finally, use LLMs for direct dialogue, summarization, and review of everything you've done on your computer - pretty impressive!

This tool serves as an open-source alternative to Rewind.ai, compatible with Ollama, and supports Chinese OCR. Interested users can deploy it locally.