SCRDeepSeek, a Chinese AI research company backed by hedge fund High-Flyer Capital Management, has released DeepSeek-R1-Lite-Preview, a new AI model designed to rival OpenAI’s o1. The model specializes in reasoning capabilities, allowing it to spend extended time considering questions before providing answers. According to DeepSeek, their model performs comparably to OpenAI’s o1 on established AI …
Microsoft announced significant updates to its AI agent ecosystem at its Ignite conference, expanding support to include 1,800 models through Azure AI. The company revealed that over 100,000 organizations are currently using or editing AI agents through its Copilot Studio platform. The updates include new autonomous capabilities that allow agents to work independently without human …
Google has launched a new memory feature for its Gemini chatbot that allows the AI to retain information about users’ preferences and characteristics across conversations. According to tech journalist Kyle Wiggers, the feature is currently available only to Google One AI Premium subscribers paying $20 monthly and works exclusively in English on the web client. …
AnyChat, a new development tool, enables seamless integration of multiple large language models (LLMs) through a single interface. Developer Ahsen Khaliq, machine learning growth lead at Gradio, created the platform to allow users to switch between models like ChatGPT, Google’s Gemini, Perplexity, Claude, and Meta’s LLaMA without being restricted to one provider, as reported by …
AI-powered search engine Perplexity has introduced a new shopping feature that enables direct purchases through its platform for Pro subscribers in the United States. As reported by Ivan Mehta, the service integrates with various sellers, including Shopify merchants, to provide product recommendations, pricing details, and one-click checkout capabilities. The feature displays visual cards with product …
Alibaba Cloud has launched an upgraded version of its Qwen2.5-Turbo AI model that can now process contexts of up to one million tokens, equivalent to approximately 1.5 million Chinese characters or 10 full-length novels. The improved model achieves 93.1 points on the RULER long text evaluation benchmark, surpassing GPT-4’s score of 91.6. According to Alibaba, …
French AI startup Mistral has unveiled Pixtral Large, a new 124-billion-parameter language model, alongside major updates to its Le Chat platform, reports Carl Franzen. The new model features advanced multimodal capabilities, including image processing and optical character recognition, while maintaining a significant context window of 128,000 tokens. The model is available for research purposes through …
Microsoft has reached an agreement with HarperCollins Publishers to use selected nonfiction books for training artificial intelligence models, as reported by Hannah Miller and Dina Bass for Bloomberg News. The deal allows Microsoft limited access to HarperCollins’ backlist titles, with authors having the option to opt out. The technology company plans to use the content …
Meta is establishing a new product group dedicated to developing AI tools for the 200 million businesses using its platforms, Sara Fischer reports for Axios. Former Salesforce AI CEO Clara Shih will lead the initiative as VP of business AI, focusing on creating and monetizing AI solutions for business users. The group aims to build …
Coca-Cola has sparked controversy with its new AI-generated Christmas advertisement that recreates imagery from its 1995 “Holidays Are Coming” campaign, NBC reports. The video, created by three AI studios using multiple generative AI models, shows animated scenes of trucks and people celebrating with Coca-Cola products. Artists and creative professionals have strongly criticized the company’s decision …