TODAY'S AI NEWS: Ilya Sutskever's SSI is taking a new route to achieving advanced AI Plus, more news from Microsoft, Reflection AI, X, Hedra, Luma Labs, Sudowrite, and World Network. Here's everything you need to know:
GENERATIVE AI
-
Manus AI: China’s new multimodal agent disrupts web automation
By
–
Cette nouvelle IA dénommée Manus fait le buzz en Chine. Et pour cause : elle crée des sites web, gère vos finances et planifie vos voyages. Une fois de plus, la Chine et les USA écrivent l’histoire. Et comme toujours, l’Europe regarde le train passer.pic.twitter.com/S1sQZYP3Ow
— Rafik Smati (@RafikSmati) 10 mars 2025Cette nouvelle IA dénommée Manus fait le buzz en Chine. Et pour cause : elle crée des sites web, gère vos finances et planifie vos voyages. Une fois de plus, la Chine et les USA écrivent l’histoire. Et comme toujours, l’Europe regarde le train passer.
-
China Launches Agent Hospital: AI-Powered Medical Platform
By
–
China's "Agent Hospital": The Rise of AI-Driven Healthcare
— Pascal Bornet (@pascal_bornet) 10 mars 2025
Tsinghua University has launched Agent Hospital, a fully AI-powered medical platform designed to diagnose and treat patients with unprecedented speed and efficiency. Staffed by 42 AI doctors across 21 specialties, the… pic.twitter.com/g59M3asffIChina's "Agent Hospital": The Rise of AI-Driven Healthcare Tsinghua University has launched Agent Hospital, a fully AI-powered medical platform designed to diagnose and treat patients with unprecedented speed and efficiency. Staffed by 42 AI doctors across 21 specialties, the
-
Creating AI Avatars for Meetings and YouTube Video Production
By
–
Finally got to know how I can make my clone or avatars for attending my long meetings and also recording some of my yt videos where I can just give my voice.
— Krish Naik (@Krishnaik06) 10 mars 2025
Specially very helpful in meeting where we can relax and get our work done 🙂
All this was possible with the help of… pic.twitter.com/ykB8PjTdpWFinally got to know how I can make my clone or avatars for attending my long meetings and also recording some of my yt videos where I can just give my voice. Specially very helpful in meeting where we can relax and get our work done 🙂
All this was possible with the help of -
AI21 Labs Prioritizes Enterprise Value in Model Development
By
–
We build and measure our models first and foremost according to the value they bring to the enterprise. Right now we see less of that in the enterprise. As these trends evolve to include additional players, we will of course update future benchmarks accordingly.
-
AI21 Labs Benchmarks Enterprise Models Against Industry Competitors
By
–
As builders of AI systems for the enterprise, we build and measure our models according to the value they bring to the enterprise. That's why our benchmarks focus on comparisons with other enterprise-focused open model providers, such as Mistral, Cohere, and Llama
-
AI Research in Mathematics: What Real Discovery Requires
By
–
Interesting thread from Daniel Litt on what it takes to do real research in math for an AI and a human I guess we had similar thoughts this week as it's really aligned with my AI-Einstein short essay: https://
thomwolf.io/blog/scientifi
c-ai.html
… -
Build Autonomous Deep Research Agent with LangGraph GPT-4
By
–
Build a Deep Research Agent An autonomous research agent built with LangGraph that combines GPT-4 and Tavily AI to perform parallel research and generate structured reports, all through sophisticated state management. Check out the implementation: https://
analyticsvidhya.com/blog/2025/02/b
uild-your-own-deep-research-agent/
… -
Deno and LangChain.js: Building AI Apps with Local LLMs
By
–
Deno Meets Local LLMs Build powerful AI applications with Deno and LangChain.js in this hands-on guide. Featuring local LLM integration, structured outputs, and visual workflows using RunnableSequence. Check out the guide on the Deno blog https://
deno.com/blog/the-dino-
llama-and-whale
… -
Hybrid RAG with EnsembleRetriever for Superior Search
By
–
Hybrid RAG with EnsembleRetriever EnsembleRetriever enhances RAG accuracy by combining semantic search and BM25 ranking in a powerful hybrid approach, delivering superior search performance. Watch now: https://
youtube.com/watch?v=UVQxMk
fQQbw
…