Reddit launches ML-powered translation, adding 35+ languages across Europe, Asia, and Latin America. Features include automatic post and comment translation, language-specific interfaces, and search indexing, enhancing accessibility for non-English users.

Allen Institute for AI launches Molmo, a series of open-source multimodal models up to 72B parameters outperforming GPT-4o. Concurrently, Meta introduces Llama 3.2 with similar multimodal capabilities at Connect 2024.

The 2024 Baidu Cloud Intelligence Conference in Beijing showcased AI advancements, including upgrades to Baige AI Platform 4.0 and Qianfan Large Model Platform 3.0. Baidu launched enhanced Code Assistant, Intelligent Customer Service, and Digital Humans. The Wenxin model reached over 700M daily invocations, with costs reduced by 90%, facilitating widespread enterprise AI adoption.

Google and Volkswagen partnered to embed generative AI into Volkswagen's myVW app for 2024 Atlas models. Utilizing Google Cloud’s Vertex AI and Gemini, the virtual assistant offers vehicle info, services, and enhanced user support, accessible via smartphones from 2025.

At Meta Connect, Meta unveiled Orion AR glasses, the $299 Quest 3S VR headset, enhanced Ray-Ban smart glasses, AI-generated content for Facebook and Instagram, new celebrity AI voices, and an upgraded Llama model with image processing.

OpenAI CTO Mira Murati resigns to pursue personal projects, along with other executives. CEO Sam Altman ensures a smooth transition amid the company’s growth and upcoming Dev Day. OpenAI continues expanding toward a for-profit model.

OpenAI restructures as a capped-profit entity, potentially valuing it over $150B. CEO Sam Altman to receive equity, shifting control from the nonprofit board. Key executives, including CTO Mira Murati, have resigned.

Meta unveils Llama 3.2 with four advanced AI models, including compact and visual-capable versions optimized for edge and mobile devices. Enhanced accessibility, multimodal features, and streamlined deployment via Llama Stack Distribution support developers and enterprises.

Arizona State researchers assessed OpenAI's o1 model with PlanBench, achieving 97.8% accuracy in standard tasks. However, performance declined in complex and randomized scenarios, and high operational costs raise questions about its practicality compared to traditional algorithms.

Broadcom introduced the Sian2 chip for AI cluster optical networks, doubling bandwidth to 200 Gbps per channel. It enhances data accuracy, reduces transceiver counts and costs, supports advanced error correction, and uses a 5nm process. Early samples are with customers.

Spotify is expanding its AI-powered personalized playlist feature to the US, Canada, Ireland, and New Zealand. Premium users can access the beta via the mobile app, creating custom playlists from text descriptions. The feature, initially launched in the UK and Australia, generates 30-song playlists and is still in testing with ongoing improvements.

James Cameron joins Stability AI's board, emphasizing AI's role in revolutionizing film production. This aligns with industry trends of studio-AI collaborations, despite legal challenges and prior AI-related criticisms.

Warner Bros. Discovery partners with Google Cloud to implement AI-generated hidden subtitles on Max. Using Vertex AI, WBD aims to reduce subtitle production time by 80% and costs by 50%, initially applying to unscripted shows with human oversight for quality assurance.

Starting Q4 2024, Google integrates Gemini AI into Workspace Business, Enterprise, and Frontline plans, removing separate purchase. Gemini protects enterprise data, offers admin controls, and has SOC and ISO certifications. Google also launches Security Advisor tool for IT admins to enhance Workspace security.

Microsoft previews the "Correction" feature in Azure AI Studio, enhancing AI accuracy by detecting and rewriting errors. It aligns outputs with benchmark documents using language models, similar to Google’s Vertex AI. Reliability is yet to be validated.

OpenAI launches advanced voice mode for ChatGPT Plus and team users, powered by GPT-4o. It offers natural, human-like conversations with real-time emotional feedback, customizable settings, and five new pronunciation options. Initially available in the US, with broader rollout planned. Enhancements include better accent recognition and a refreshed interface.

Intel's Altera launched Agilex 3 and 5 FPGAs with enhanced AI capabilities for edge and cloud computing. New hardware, software, and tools expand programmable solutions across industries. Altera aims to lead the global FPGA market with improved performance, security, and developer support.

Anthropic seeks a $40B valuation in its new funding round, supported by Amazon and major tech firms. It will present at the Morgan Stanley conference. Competitor OpenAI recently raised $6.5B at $150B. Anthropic has secured $7.6B, advancing AI with its Claude model and strategic partnerships.

OpenAI launches MMMLU dataset in 14 languages to enhance global AI. Using expert translations, it expands the MMLU benchmark. OpenAI Academy supports AI education in emerging markets, advancing multilingual AI for businesses.

OpenAI launches OpenAI Academy to foster global AI innovation, focusing on low and middle-income countries. Participants gain training, $1M API credits, and resources to develop AI solutions in sectors like healthcare and education.