Bloomberg Línea — Microsoft’s move to hire Sam Altman following his dismissal from OpenAI attracted the spotlight of the generative artificial intelligence industry last week, before the tech entrepreneur returned - until further notice - to the helm of the startup considered the most disruptive in this market. But for Microsoft, this hasn’t changed the plans to advance and monetize the technology later this year.
“We have made the OpenAI service available within Edge [Microsoft’s browser] for the entire Brazilian market, allowing us to work with low latency solutions to serve the public and the companies that are installed here in Brazil,” Tânia Cosentino, Microsoft’s CEO for Brazil, said in an interview with Bloomberg Línea prior to the incidents involving OpenAI’s CEO.
Cosentino stressed that Microsoft “is already seeing great returns on investment in the area of software development with AI”, without revealing specific figures.
And Microsoft is not alone. Big tech companies such as Google, Meta and Amazon, to name just a few, have also stepped up the pace of launches in recent months.
For Microsoft, one of its biggest bets is on its ambitions to change forever the computers that run the Windows 11 operating system.
At the end of October, Microsoft announced what promises to be “a new era of AI” in the company, with Microsoft Copilot, as well as updates for Windows 11, Microsoft 365, Bing, Edge and Surface, all focused on expanding AI capabilities.
Microsoft, which is the largest investor in OpenAI, the startup that developed ChatGPT, will now have Copilot - which was developed by GitHub, a platform for developers bought in 2018 - exclusively incorporating the context and intelligence of the web and the working data of what the user is doing on the computer to provide higher quality assistance.
The generative AI assistant, available in Windows 11 since the end of October, works like an app and can “pop up” when the user needs it. With more than 150 new features, this update is one of the most ambitious to date, according to Microsoft, as it extends the power of Copilot and enables new AI-powered experiences for applications such as Paint, Photos and Clipchamp.
Cosentino said Microsoft has invested heavily in local infrastructure, such as data centers, to expand services and provide low-latency AI solutions to the Brazilian market. She says the demand for qualified IT professionals is evident both in Brazil and worldwide, which makes generative AI a valuable tool for increasing developer productivity, with calculated gains of more than 50%.
“Through voice commands, the system develops the software for you or tests your software, debugs it, and this brings productivity gains of more than 50% within the developer area. This is an immediate gain for companies,” she told Bloomberg Línea.
Since the beginning of November, Microsoft 365 has made Copilot available to commercial customers, with an updated version of M365 Chat and new features in applications such as Outlook, Excel, Loop, OneNote, OneDrive and Word.
Bing Chat Enterprise will also receive updates, including support for multimodal visual search and the availability of Image Builder in the Microsoft Edge mobile app - all supported by generative AI.
Potential for corporate solutions
“The world of technology is permeating every company. IT is no longer just a support area to keep the operation ‘running’ and is now a strategic area for business development, for solving new products, for interacting with new customers and for productivity gains and cost reduction in general”, said Cosentino.
She also said that Microsoft is testing various bots and is noticing not only greater precision in answering users’ questions, but also a much shorter response time.
“This means transforming the user experience in relation to a support area. You improve the experience, user satisfaction, brand loyalty and, at the same time, reduce the entire cost of interaction for your teams,” she said.
Copilot’s development clients in Brazil include companies such as Sem Parar and public bodies such as the Federal Attorney General’s Office and the Federal Court of Auditors.
“These are companies with large volumes of documents and processes. It becomes possible to equalize these documents, create them, classify them and seek answers for different processes, all with the use of generative AI,” said Cosentino. “These are cases we already have under way in Brazil and around the world, applications with a guaranteed return on investment,” she said.
The market vision
Leonardo Otero, founding partner of Arbor Capital, which specializes in technology, said that the launch of Microsoft’s Copilot in Windows 11 will completely transform the way users interact with computers. He predicted that the combination of Copilot and Office365 will result in a significant increase in productivity and said that soon no competitive professional will be able to do without a generative AI assistant.
Otero said that he believes the price of this package will be substantial, but with most of the value representing profit, since the variable costs are low.
And he pointed out that Copilot will be especially useful for users who rely heavily on email, with the summarization of long texts and generation of replies that can be passed off as authentic. Productivity gains have a positive impact, not only on individuals but on the economy as a whole, in his assessment.
This advance in generative AI is so significant that, according to Otero, Microsoft may even find an opportunity to compete in the cell phone market with Apple.
But there are various players already pursuing these gains at an advanced stage.
Meta has created an AI model called SeamlessM4T, which can translate and transcribe almost 100 languages into text and speech and has the ability to convert audio into text.
In addition, OpenAI also announced at the end of October that ChatGPT can now respond out loud, and Getty Images launched a new image creation tool based on generative AI.
Spotify announced that a pilot group of podcasters can now automatically translate their podcast episodes into alternative languages while maintaining their own voice. The feature uses AI to translate podcasts into Spanish, French and German.
Google also announced features for its generative AI, Bard, which interacts with other Google applications and services, with personalized responses for each user.
Among the new features is the possibility of uploading images with Google Lens, receiving images as a response and modifying the texts received by Bard, making them simpler or more complex, shorter, longer or more casual or professional, according to the CEO of Google in Brazil, Fábio Coelho, in a post on his LinkedIn page.
-- Translated from the Portuguese by Adam Critchley