What's in Tech : Week of November 10th 2024

What's in Tech : Week of November 10th 2024

In this week’s "What’s in Tech" newsletter, we highlight the latest developments in the AI space. Key updates include the launch of Microsoft’s new AI agent platform, OpenAI’s new feature to improve LLM latency and Google’s new AI video creation tool. We also saw executive movements at OpenAI and AWS, and partnerships between Anthropic and Palantir; and Infosys and Inspeq AI.

?

Centific at LOC World 52 in Monterey

?We demonstrated Flow, our LLM-powered application for direct AI translation, built on our AI Data Foundry platform. The application helps organizations automate and streamline complex localization workflows with flexibility, security, and precision.

?Additionally, Rimi Endo, Senior Director – Global Client Partner, showcased the power of AI in localization alongside Amazon Music. Their co-presentation, titled “Advanced Localization Management”, highlighted innovative content and processes, emphasizing the importance of industry leadership in driving advancements.

?

Microsoft Launches an Open-Source Multi-Agent AI Platform

Microsoft unveiled a new multi-agent infrastructure called Magentic-One that allows a single AI model to power various helper agents that work together to complete complex, multi-step tasks in different scenarios.

OpenAI Hires Meta’s Former Orion Head to Lead its Robotics Efforts

The former head of Meta’s Orion augmented reality glasses initiative Caitlin Kalinowski has joined OpenAI to lead the startup’s robotics and consumer hardware efforts.

OpenAI Introduces ‘Predicted Outputs’ Feature for Improving Latency in LLMs

The key innovation lies in the ability to predict probable content and use it as a starting point for the model, effectively skipping portions of the process where the outcome is already well-established. Also, latency can be decreased by as much as fivefold, making GPT-4o far more suitable for real-time tasks.

Meta Extends Llama Support To U.S. Government For National Security

Meta has just announced the availability of its open-source Llama models to U.S. government agencies in the areas of defense and national security, along with their private-sector partners.

Amazon May up its Investment in AI Startup Anthropic

Amazon is in talks to invest multiple billions in Anthropic, but reportedly wants the startup to use Amazon-developed silicon hosted on Amazon Web Services to train its AI instead of NVIDIA.

Anthropic and Palantir Partner to Bring AI Models to AWS for US Government Intelligence

The partnership facilitates the responsible application of AI, enabling the use of Claude within Palantir’s products to support government operations such as processing vast amounts of complex data rapidly, elevating data driven insights, identifying patterns and trends more effectively.

Sandeep Dutta Appointed as President of AWS India and South Asia

He has worked in a variety of roles at Accenture for more than 16 years, most recently as the Senior Managing Director and India Business Lead.

Google's New AI Feature Can Create Marketing Videos

Google has launched video maker Vids across multiple Workspace editions. The tool uses the Gemini AI model to create workplace and marketing videos from Google Drive files and descriptions.

HCLTech Partners with Inspeq AI to Promote Responsible AI in Global Enterprises

This partnership focuses on integrating Responsible AI practices into existing tools like Microsoft Co-Pilot and AWS Bedrock. The collaboration aims to improve processes in various industries, particularly in financial services.

Tencent Releases Open-Source MoE LLM Hunyuan-large

The model has a total of 398 billion parameters, making it the largest in the industry, with 52 billion activation parameters. Public evaluation results show that it leads in various disciplines and tasks including NLP in Chinese and English languages, code processing, and mathematics.

Hugging Face and SambaNova Create One-Click Chatbot Integration for Developers

The integration supports both text-only and multimodal chatbots, capable of processing both text and images. Developers can access powerful models like Llama 3.2-11B-Vision-Instruct through SambaNova’s cloud platform, with performance metrics showing processing speeds of up to 358 tokens per second on unconstrained hardware.

AI4Bharat and IBM Research India Release New Indic Language LLM Benchmark

This benchmark, developed under The AI Alliance, includes 85,000 multiple-choice questions across 11 Indian languages, covering eight diverse domains and over 40 subjects with an India-centric focus on both general and cultural knowledges.

Mistral Launches a Content Moderation API

The is ?powered by a fine-tuned model (Ministral 8B) trained to classify text in a range of languages, including English, French, and German, into one of nine categories: sexual, hate and discrimination, violence and threats, dangerous and criminal content, self-harm, health, financial, law, and personally identifiable information.

Regards

Dinesh Chandrasekar

?

Julio Pinet

CEO and Founder @ Global Initiative | Spearheading Global AI Collaboration

2 周

What incredible insights you have shared with us, Dr.Dinesh Chandrasekar (DC)regarding the latest developments in the AI space. I am particularly excited about Microsoft's multi-agent, Magentic-One, I can't wait what it brings to the industry. Thanks again.

要查看或添加评论,请登录