Microsoft aproveita IA no Microsoft 365 Copilot

Microsoft leverages AI in Microsoft 365 Copilot

The way we work is changing rapidly as mass adoption of AI-based productivity tools is upon us. Here's what you need to know about Microsoft 365 Copilot.

Imagem em destaque

In the race for AI-powered productivity, Microsoft announced and launched Microsoft 365 Copilot, a tool that promises companies to transform the way we work, supposedly unleashing creativity, unlocking productivity and improving skills. This was first triggered by the emergence of generative AI which became a trending innovation topic after San Francisco-based startup OpenAI introduced the now world-famous ChatGPT chatbot in November 2021. Six months later, ChatGPT is expected to launch its fifth iteration in early 2024 .

This context of in-person innovation serves as the backdrop for many tools that leverage OpenAI, such as Microsoft 365 Copilot. It is public knowledge that both companies have collaborated over the past four years in hopes of accelerating advances in AI. In January 2023 both parties declared that they would extend their long-term partnership with Microsoft for a multibillion-dollar investment . Microsoft and OpenAI have joined forces to develop the underlying AI technology. While Microsoft is responsible for building and delivering the supercomputing infrastructure, OpenAI operates its services on this infrastructure. Companies at all levels and in all sectors will certainly benefit from this collaboration. But how, exactly?

How Microsoft 365 Copilot works

Microsoft Copilot is built using the Azure OpenAI Service , a Microsoft service that provides custom access to the OpenAI API. Azure, as it is commonly known, is a cloud computing platform run by Microsoft, different from Azure OpenAI. As Microsoft explains , the latter is used to provide customers with rich language AI with OpenAI GPT-4, GPT-3, Codex, and DALL-E models with the security and enterprise promise of Azure. This should not raise enterprise data security concerns, as data sent to Azure OpenAI services is not passed to OpenAI for this company's model predictions. The data remains in Microsoft Azure, with exclusive control and governance of the data.

This Copilot is a highly sophisticated tool that harnesses the power of large language models (LLMs) to increase productivity and streamline workplace workflows. Using state-of-the-art GPT-4, Copilot has the ability to generate professional-grade letters and documents, provide comprehensive summaries of complex data, create visually appealing PowerPoint presentations, and perform in-depth analysis of Excel spreadsheets. Its advanced features include intelligent meeting outline and summary, robust spreadsheet analysis, and the ability to select relevant graphics to complement presentation content. In short, practically all of a company's stakeholders will benefit from this tool in one way or another.

Integration into Microsoft 365

Copilot is integrated with Microsoft 365 in two ways. The first is built into popular applications such as Word, Excel and PowerPoint. The second is through an innovative new feature called Business Chat, which leverages LLM and its data to generate natural language prompts and status updates. This empowers stakeholders to drive better business results and perhaps this is the feature that will set Copilot apart, at least until competitors like Google launch a parallel product, but more on that later. Let's explore the most popular apps and how Copilot elevates each of them.

Copilot in Word

Microsoft's promise is to be able to start the creative process without fearing the blank page. Copilot will render a first draft for editing and tweaking, saving you time in drafting, sourcing, editing, and reviewing. This doesn't mean the content will be trustworthy all the time. This is where valuable human oversight and intervention will shine. Copilot's job will be to put you further ahead, shaving days or even weeks off a traditional creative schedule. People will always be in control as authors, but it will be easier to shape ideas and messages by asking the Copilot to summarize, rephrase, or provide feedback.

Copilot in PowerPoint

It will help you create professional presentations with simple instructions, taking relevant information from your own documents. Slideshows will be assembled in seconds with as much information as needed and in the correct visual style. Think about the many hours you or your colleagues, as non-design professionals, have spent trying to get your branding right, playing with the size of text boxes or choosing the right graphics to accommodate data. Refining the results can be done manually or even with the help of Copilot. In other words, you can train it to function the way you do.

Copilot in Excel

Analyzing data, identifying trends, and finding new insights will be a game changer for those who aren't Excel experts or simply don't have enough time. Copilot will be able to collect relevant data and create professional visualizations for those time-consuming periodic reports. Antonio Bartolo, Microsoft's chief cloud advocate suggests that most people only use 10% of Excel's capabilities and that Copilot can help them leverage the full potential of this Office application. Helping users unlock the full benefits of an application can be the case for Excel and any other.

Copilot in Outlook

Hundreds or thousands of unread messages in your inbox won't cause anxiety. By opening a chat window where you can instruct Copilot using natural language, you can clear your inbox in minutes, respond to emails with suggested responses, and like other text generators powered by AI, improving tone, clarity or conciseness will be part of the basic functionalities. You can also request feedback on an email you wrote that exemplifies how you can improve your skills using Copilot.

Copilot in teams

We all know the modern saying: That meeting could have been an email. Responding to this sentiment, Copilot can summarize key points in a minute, specifying who said what and where people agreed and disagreed, all in real time. The immediate benefits of this will be faster decision-making in team collaboration and the definition of next steps – all written down and ready to be executed when the meeting ends.

Business chat

This could be the crown jewel of the Copilot. Business Chat allows you to use your natural language to request information, i.e. through prompts. It works across all Microsoft 365 apps and will respond to your request depending on the app you're getting information from and the context or intent of your question. Regardless, it will help surface relevant insights from an abundance of data.

What to expect

Microsoft Copilot was publicly announced on March 16, 2023. The launch saw limited distribution among a small selection of customers, and the tech giant plans to launch the full app in late 2023. By then, some companies will have tried it, liked it benefits and tested its limitations. The first productivity insights could show us the true potential of this tool.

For example, we know Github Copilot AI Peer Programmer was a revolutionary product for developers, not just because it increased individual developers' productivity, but because it can leverage an organization's untapped data and insights. Github has also documented its impact in at least two aspects . The first is improving developer satisfaction, with 60–75% of users reporting that they feel more satisfied with their work, and the second is conserving mental energy, with 73% reporting that it helped them preserve mental effort during repetitive tasks . Its latest release, Copiloto X , was announced in March 2023, featuring, as the website explains, chat and terminal interfaces, support for pull requests, and early adoption of OpenAI's GPT-4. Copilot X will be able to suggest code improvements, explain errors and even generate unit tests.

This is truly how AI-based solutions serve us: the opportunity to focus on high-value tasks that contribute to the purpose of our work. On the other hand, as Microsoft 365 Copilot continues to develop, common concerns are emerging not only about data security, but also ethical, legal, and social issues. The reality is that AI has just gone mainstream. It is natural to ask ourselves about the consequences of our trust in this technology and whether we can guarantee its responsible use. Responsible AI aims to build trust, ensure accountability, and ensure accessibility. As such, it is a fundamental commitment for Microsoft and its future developments. Responsible AI is not just a concept promoted by the World Economic Forum, it is also expected to become standard practice for the governance of AI in the developments and innovations to come.

When it comes to technology, we've learned to always expect the unexpected. At the time of this writing, Google is working on developing Google AI for Workspace , a product similar to Copilot that will work in Docs, Sheets, Slides, and other Workspace apps. While there is no release date for this solution, we are excited to learn about the new ways AI and continuous innovation are shaping the future.

If you liked this, be sure to check out our other articles on AI.

  • Modern algorithms that will revolutionize your business
  • Move Over Stack Overflow – ChatGPT wants to take the crown
  • Neuromorphic Computing: Discover the Future of AI
  • Not adopting AI? You could be hurting your productivity
  • Unmatched Accuracy: Optimizing Business Analytics with AI and BI

Source: BairesDev

Conteúdo Relacionado

Vídeos deep fake ao vivo cada vez mais sofisticados...
Aprenda como os processos baseados em IA aprimoram o...
O Rails 8 sempre foi um divisor de águas...
A GenAI está transformando a força de trabalho com...
A otimização de processos industriais é um desafio constante...
A Microsoft acaba de apresentar sua nova biblioteca Microsoft.Extensions.VectorData.Abstractions,...
Entenda o papel fundamental dos testes unitários na validação...
Aprenda como os testes de carga garantem que seu...
Aprofunde-se nas funções complementares dos testes positivos e negativos...
Entenda a metodologia por trás dos testes de estresse...
Descubra a imprevisibilidade dos testes ad hoc e seu...
A nomeação de Nacho De Marco para o Fast...
O mercado embarcado tem uma necessidade de soluções de...
A Inteligência Artificial (IA) tem se tornado cada vez...
Ao relatar estatísticas resumidas para resultados de testes de...
Como você previne alucinações de grandes modelos de linguagem...
Nos últimos anos, a Inteligência Artificial Generativa (Generative AI)...
Domain-Driven Design (DDD) é uma abordagem estratégica importante para...
Back to blog

Leave a comment

Please note, comments need to be approved before they are published.