Microsoft today announced that Sebastien Bubeck, its vice president of generative AI research, is departing to take a comparable job at OpenAI, the AI startup behind ChatGPT.

Details about Bubeck's new position at OpenAI were not shared but a Microsoft spokesperson said he is leaving, stating, "Sebastien has decided to leave Microsoft to continue his efforts in the pursuit of Artificial General Intelligence (AGI)." Sebastien Bubeck A spokesperson made sure to note that the company would be looking forward to continued collaboration through his work at OpenAI, which is backed by Microsoft.

Bubeck didn't comment on a request. His resignation comes after a couple of high-profile departures from OpenAI over the past few months. Among them is former CTO Mira Murati, who left in September. Even with this series of changes in personnel, the company's restructuring efforts had nothing to do with the actions, according to OpenAI Chief Executive Officer Sam Altman.

Another significant change occurred in the AI research landscape as experts like Bubeck move from one leader to another in AI research organizations.

---------------------------------------------------------------------------------------------------

For More Details About Full Stack Courses : Visit Here
Register For Free Demo on UpComing Batches : Click Here

"Meta’s new Llama 3.2 model, available in 11-billion and 90-billion parameter variants, can process images and text. It understands charts, captions images, and locates objects from natural language prompts."

Meta has launched its latest AI model, Llama 3.2, marking a significant step in the open-source AI space. This release introduces models with both image and text processing abilities, following two months after their last major AI update.

Llama 3.2 comes in several variants, including 11-billion and 90-billion parameter models that are designed for advanced visual tasks. These vision models can read charts, caption images, and identify objects using natural language commands. The larger, 90-billion parameter model is even more advanced, capable of extracting details from images to create precise captions. For developers looking for AI solutions that handle text, there are also smaller models with 1-billion and 3-billion parameters, designed to work efficiently on mobile devices and edge hardware.

These models are particularly useful for developers creating AI-powered applications, such as augmented reality apps that can understand videos in real-time, visual search engines, and document analysis tools. The text-only models are optimized for devices running on Qualcomm and MediaTek hardware, enabling features like summarizing conversations, scheduling meetings, and building personalized AI apps.

Although Meta is still catching up with competitors like Anthropic and OpenAI in the multimodal AI space, Llama 3.2 offers strong competition. Meta claims that Llama 3.2 performs better than Gemma and Phi 3.5-mini in tasks such as rewriting prompts, following instructions, and summarizing content.

This new release reinforces Meta’s commitment to pushing forward in AI development, providing versatile tools for developers to build next-gen AI applications.

----------------------------------------------------------------------------------------------------

For More Details About Full Stack Courses : Visit Here
Register For Free Demo on UpComing Batches : Click Here

Amazon Web Services (AWS) has been chosen to provide cloud computing services to Indian public sector banks through the Public Sector Bank Alliance (PSBA).

Amazon Web Services (AWS) has been selected by the Public Sector Bank Alliance (PSBA) to provide cloud computing services to public sector banks across India. The PSBA, which includes 12 major public sector banks, acts as a central body to help these banks implement advanced, cloud-based banking services.

Some of the banks in this alliance include prominent names like State Bank of India (SBI), Punjab National Bank (PNB), Union Bank of India, Canara Bank, and Bank of Baroda, among others.

According to AWS, this selection will allow these banks to easily adopt AWS cloud services without having to go through individual procurement processes. AWS will deliver these services via its Managed Service Providers (MSPs), Orient Technologies and Hitachi Systems.

At the Empower India event, Amazon CTO Werner Vogels highlighted the importance of cloud solutions in India. He noted that businesses of all sizes, including startups and large enterprises, are focused on maximizing the benefits of cloud services. This is particularly important in India, where scalable and efficient digital infrastructure is needed to serve a population of over 1.4 billion people.

Besides the banking sector, AWS is also working with other government bodies like the National Health Authority (NHA) and Government e-Marketplace (GeM) to enhance digital services. NHA is leveraging AWS for its flagship Ayushman Bharat-Pradhan Mantri Jan Arogya Yojana, a healthcare program aimed at improving health services on a large scale. AWS solutions are being used to manage everything from beneficiary identification to claim processing in a secure and scalable manner.

GeM, India’s online government procurement platform, is using AWS cloud services to power various functions, including direct purchasing, reverse auctions, and price matching.

AWS India’s public sector leader, Pankaj Gupta, emphasized that India’s public sector is playing a crucial role in driving the country’s digital transformation. He stated that the cloud and AI technology are helping the country work toward the government’s vision of a “Viksit Bharat 2047”—a roadmap to a developed India.

AWS has also been supporting several other government initiatives, such as the National Skill Development Corporation (NSDC), Swachh Bharat Mission, Co-WIN, and DigiLocker, highlighting its role in the country’s digital journey.

----------------------------------------------------------------------------------------------------

For More Details About Full Stack Courses : Visit Here
Register For Free Demo on UpComing Batches : Click Here

Last month, the 2024 edition of the .NET Conf: Focus series took place, concentrating on AI development within the .NET framework. This event catered to developers of all experience levels, offering both informative and interactive sessions that showcased how AI can be integrated into .NET applications. The conference stood out as one of the most technical events in the Focus series, with numerous in-depth presentations.

The Focus series is a spin-off from the larger annual .NET Conf, a free event organized by Microsoft and the .NET community. Each Focus event highlights a particular technology related to .NET, featuring a variety of in-depth, hands-on sessions for developers. These events are one-day, live-streamed events with talks from community experts and members of the .NET product teams.

.NET Conf: Focus on AI was unique, as it centered on the latest AI developments and how they intersect with the .NET ecosystem. The sessions covered a wide range of topics, from introductory AI concepts like Retrieval-Augmented Generation (RAG) to advanced technical discussions on AI tools in .NET applications.

The conference opened with a keynote by Scott Hanselman, Microsoft's Vice President of Developer Community, and Maria Naggaga, Senior Program Manager for Visual Studio and .NET. They emphasized that AI is a fundamental part of the future of .NET, highlighting how modern applications are evolving into intelligent systems.

During the keynote, several AI concepts were introduced, including ethics in AI and RAG, a topic of particular interest to developers unfamiliar with AI. Naggaga explained how RAG could be effective, particularly in customer service scenarios where AI works with real-time data. Hanselman demonstrated GitHub Copilot as an intelligent coding assistant, emphasizing the use of domain-specific data to create smarter applications.

The second session was led by Stephen Toub, a Distinguished Engineer at Microsoft, who demonstrated how to incorporate AI into existing .NET apps. Toub live-coded a console app, showing how Semantic Kernel could be used to simplify AI integration by separating AI services from the core logic of the application.

Daniel Roth, Program Manager for ASP.NET, conducted a hands-on session on creating AI-powered web apps using Blazor and .NET, showing how to leverage AI for enhanced user experiences. Bruno Capuano, an AI and .NET Advocate at Microsoft, presented a session on integrating AI models into .NET applications, using both local and remote small language models like Microsoft’s Phi-3 for text and vision processing, which the audience found valuable.

Several shorter sessions followed, most lasting around 30 minutes, and continued to explore topics introduced in the keynote. While many sessions focused on Azure-based solutions, they provided actionable insights for building AI-powered applications in .NET. Other notable talks included the use of AI agents to automate workflows, building personalized AI assistants using Teams AI, and applying vector embeddings in .NET for recommendation systems and RAG applications.

A key takeaway from the event was the audience’s concern about privacy when using AI, particularly with RAG, which relies on domain-specific data. Although there was no dedicated session on this, engineers addressed some of these concerns during live Q&A and mentioned plans to publish a technical article on the subject. Another prominent concern was the cost of using Azure services, a recurring issue based on past cloud projects. Additionally, attendees questioned the future of ML.NET, with some wondering if Microsoft would discontinue it in favor of newer AI tools.

For more details, the full agenda is available online, and the entire event can be watched on YouTube. Session slides and demo code can be accessed via GitHub.

----------------------------------------------------------------------------------------------------

For More Details About Full Stack Courses : Visit Here
Register For Free Demo on UpComing Batches : Click Here