OpenAI is the pioneer of the artificial intelligence chatbot known as ChatGPT, which is dedicated to the research and implementation of artificial intelligence (AI), referring to highly autonomous systems that can be applied to any task that involves understanding, generating code or creating new language models. As such, ChatGPT has the ability to create long and consistent responses, which do not appear to have been computer-generated, being the fundamental foundation of why this particular application is gaining momentum in the artificial intelligence (AI) industry, with a whole range of possible applications.
Microsoft will deploy this technology in several products on the market for homes and businesses and will include new categories of digital experiences, based on AI and aims to provide the necessary support for users to build cutting-edge applications. This includes the Azure OpenAI Service, which allows developers to build higher-tech solutions by using OpenAI models with the confidence and assurance that Azure offers.
Azure OpenAI Service runs on Microsoft Azure’s AI-optimized infrastructure and delivers enterprise-grade functionality, based on combining the best AI features with the best cloud capabilities. To provide organizations that need scalable, high-performance AI solutions.
The growing development of large-scale language models has made it possible to solve complex problems through the development of AI algorithms. As such, Azure is currently ranked as the only public cloud in the world, offering AI supercomputers with high scalability, because it combines networking and GPU solutions, in an unparalleled architectural design, which guarantees high performance for AI training workloads, which generate high processing consumptions.
The main goal of Microsoft is to be able to offer a complete service of AI, through which companies can make use of the most advanced intelligence models currently, but for this Microsoft will generate awareness in the responsible use of innovations in AI, so it has committed to promoting principles that put people first, since the models that are available in Azure OpenAI, although it is true have important benefits but, without a detailed design and previously analyzed mitigations, these models have the potential to generate incorrect and even harmful content, so to enter this service, Microsoft has implemented its own security measures.
One of them is limited access, which requires developer users to request access specifying what use they will make of these tools, before they can access them.
The next step is to add responsible AI content filters, which will analyze the context, and if the content is identified as abusive, hateful, or offensive, the developer will be notified to take action such as, modify, or remove this content that is categorized as inappropriate.
It is expected that several companies will be able to request access to the most advanced AI models that exist today, including the GPT-3.5 language system and Codex, as well as the DALL-E model, which is based on the generation of images from text indications, these models are backed by their high levels of enterprise-level reliability and AI-optimized infrastructure to create state-of-the-art applications.
Blog by: Michael Abad
