Oxford researchers replicate cerebral cortex structure with 3D-printed stem cells for brain injury treatment. Customized implant therapy progress made in mice. Future focus on complex multilayered brain tissue development.

Microsoft to launch AI chip codenamed "Athena" to reduce reliance on NVIDIA GPUs, compete with H100. Aim to match performance, cut costs, & attract users like AWS & Google. OpenAI may also develop own chips. Cloud providers look to diversify & reduce vendor dependence.

UNESCO partners with NCCA to launch AI supervision project for EU countries, ensuring compliance with regulations and ethics. The project aims to provide tools, knowledge, and training for safer AI deployment in European nations.

In the field of generative AI, Docker containers are key for deployment. Docker launches GenAI Stack, integrating Neo4j, LangChain, and Ollama for developers. Docker AI simplifies AI app development combining various components in a container. Docker also introduces Docker AI as a "mech suit" for developers with unique data-trained assistance for troubleshooting.

Google's October event focused on new Pixel hardware and the integration of generative AI into Google Assistant. Bard, the AI, will enhance Assistant's capabilities, but specific launch dates are unclear. Google aims to improve Assistant's performance on complex queries with Bard's natural language understanding, a potential turning point for the assistant's development. Assistant's advancements may be restricted to Google's hardware initially, with plans for testing among select users.

Oracle is revolutionizing healthcare with innovative solutions focusing on patient care, clinical expertise optimization, and cost efficiency. Their cloud-based EHR platform and generative AI services aim to streamline processes for providers and enhance patient engagement. The acquisition of Cerner reflects Oracle's commitment to the healthcare sector, aiming to improve patient outcomes and establish a significant industry position.

Meta introduced Llama 2 Long, an extended AI model outperforming competitors like OpenAI's GPT-3.5 Turbo. By modifying RoPE encoding, it generates accurate replies with less storage space. Powered by RLHF and Llama 2 chat data, it excels in language tasks and user queries, proving Meta's open-source AI can rival closed-source models.

Google's AI Bard will soon remember details to personalize responses, like location or dietary preferences. A new "Memory" feature will remember key details for future conversations. The feature can be controlled through a toggle switch for privacy concerns.

Google introduces Google-Extended, a tool for publishers to prevent AI model training using their data while allowing Google Search indexing. Websites can manage access to their content through a toggle button and robots.txt file. Some sites, like NYT, are legally blocking Google for AI training.

Researchers at Carnegie Mellon University collaborated with Adobe and UC Berkeley to create algorithms addressing copyright and compensation in generative AI models. New papers will be presented at a computer vision conference in 2023, focusing on removing copyrighted content and fairly compensating data contributors. This research aims to provide more control and freedom for artists and data owners in the AI generation process.

Cohere, an AI startup, releases chat API integrated with Retrieval-Augmented Generation (RAG) for developers to create reliable AI conversational products. RAG enhances accuracy by connecting models with external data sources like web searches and text documents, ensuring relevant, verifiable replies. Developers can use RAG modules or the end-to-end solution, enhancing transparency and credibility in content generation.

Generative AI market growing at 27% CAGR, set to exceed $22 billion by 2025. AI-human synergy seen as key for business growth, job security, and innovation. AI to enhance operations, not replace humans. Startups racing to innovate with AI, attracting investor interest.

Vise, a fintech company, is launching Vise Intelligence, an AI service to assist human financial advisors in managing client investment portfolios better. The service aims to provide personalized insights based on market data and client information, ultimately helping advisors serve more clients effectively.

Researchers have developed Temporal Graph Neural Networks (TGNNs) for dynamic graphs, outperforming static GNNs. Memory-based Temporal Graph Neural Networks (M-TGNNs) store node-level memory vectors, but face scalability challenges. DistTGL improves M-TGNN training on distributed GPU clusters, enhancing accuracy and convergence speed.

Apple is reportedly developing an internal search engine called "Pegasus" to enhance its Spotlight feature and compete with Google. Despite executive statements suggesting otherwise, Apple's robust advertising business could support the launch of its own search engine.

Visa launched a $100 million investment program in generative AI, focusing on companies developing innovative AI technologies. The program aims to shape the future of business and payments. Visa Ventures will lead the investment, supporting disruptive startups in AI and payment sectors. AI tools, like fraud prevention, are transforming CFO operations with enhanced analytical capabilities.

OpenAI's ChatGPT now includes visual and audio capabilities. Google's Gemini is set to launch this fall. Stability AI and Meta are also developing multimodal systems. Stability AI has plans to release an open-source model, while Meta is working on Llama 3 and ImageBind. The future of AI lies in multimodal systems, with various open-source initiatives already in progress.

ChatGPT's popularity highlights the expanding role of AI. Researchers reveal insights on optimizing neural networks for future AI applications. Balance between network depth and width is key for powerful and dynamic AI systems.

OpenAI's ChatGPT now utilizes "Browsing with Bing" to provide real-time info from current sources. Microsoft's Bing Chat and Google's Bard offer similar features. Meta also incorporates Bing for real-time web results. To access this feature, follow OpenAI's instructions in the iOS application settings.

Adobe released Photoshop Web Edition, an online version of their photo editing app with popular features like content-aware tools. Users can collaborate and access tutorials, but it's not free, starting at $9.99/month. New features from the desktop version will be added soon.