From Basic to Artificial Intelligence: half a century of Microsoft

Right at the peak of success, in 1995 with the release of the new improved version of Windows that was finally acceptably comparable to the Macintosh, Gates did not see the arrival of the new phase until the end of the year: the Internet was taking off. The browser manufacturer, Netscape, went public in August and obtained staggering valuations for a company that invoiced almost nothing, spent money to produce software and gave away its product on the net. But it conquered 90 per cent of the market for browsers that were bidding to be a kind of operating system of the web. One employee, according to a journalistic reconstruction, asked Bill Gates if he could distribute a browser for free to counter Netscape. Gates replied: "Give away software? Have you gone communist?". After all, in the Windows 95 architecture, the Internet was still considered a peripheral of the PC. At the end of that year, however, Gates realised the situation. And he declared that he would change course, concentrating the entire company on the web and adding the Explorer browser to Windows. This guaranteed another five years of success and brought about the end of Netscape, which could not withstand competition from its more powerful opponent, which was also investigated by the antitrust authorities. Companies, large or small, that had a strategy based on the Internet flourished and conquered the imagination about the future: Aol, Sun, Amazon, Yahoo!, Google. Gates was the richest man in the world. But he did not save Microsoft from its own prophecy: by the beginning of the 2000s, it was no longer the leader in IT.

Interviewed on this point, the new CEO Steve Ballmer, long Gates' right-hand man, said that the competition is called Google. Gates had left the position of ceo to become a 'software architect' for Microsoft before becoming a philanthropist. He perhaps felt that the times were too different from his own. What did leadership mean in a world where the network was more important than PCs? The complexity of the relationships between digital objects prevailed over the importance of the technology at the nodes. PCs were no longer the centre but one of many things that connected to the network. And digital mobile phones were becoming increasingly important. The Europeans, from Nokia and Ericsson, controlled the market and wrote their own operating systems, in turn trying to connect them to the Internet. The new order was to come in 2007 from a company that only ten years earlier had practically gone bankrupt. Steve Jobs presented the iPhone and changed the balance. Smartphones became the tools of choice for going mobile on the Internet, PCs were left on desks but no longer dictated the law, the large servers on which the programmes that could serve all terminals ran became strategic. The system, in a short time, moved towards an architecture made up of many objects connected to the network and large data centres serving them.

Amazon, one of the most important companies to have survived the dot-com bubble, had changed its structure: from an online shop it had become an e-commerce platform and invested so much in the servers on which it relied that it could also sell its computing capacity to third parties. Cloud computing became one of the most important forms taken by computing in the 21st century. Google was in fact along the same lines and competed with Office with products that were used directly on the net. What would Microsoft do?

In 2014, Satya Nadella, an Indian manager with great vision, took over from Steve Ballmer and articulated a new strategy. The Microsoft that had lost its way saw a new light. PC operating systems were still a business and Office, the suite of three traditional products, for presentations, for computing, for writing, was still important in companies, which changed habits more slowly than consumers did. But, without the centrality of PCs, they were gradually losing value. Over the next ten years, Nadella managed to reposition his products, which became cloud-based services. Office worked by connecting the PC or phone to the network and using programmes through servers that served to store the work done, share it with others, update it in a synchronised way, and so on. It was not an invention: but the strategic reorientation of a giant like Microsoft was a masterpiece. The new architecture decided by Nadella paved the way for a completely new relationship with customers. Microsoft could now collaborate on business projects with greater ease and participate more closely in business innovation, without the aura of monopolism that Microsoft could not help but carry in the previous phase. A number of acquisitions of online services, from LinkedIn to GitHub, broadened the scope of activities. And all this made it easier to eventually introduce a pre-packaged form of generative artificial intelligence, produced by OpenAI, but integrated into the enterprise software suite, looking for a way to use it efficiently for business. Copilot had entered the game. A new game for IT had begun. And the market was giving Microsoft credit like it had not in decades.

Because Microsoft showed that it wanted to ride the new wave with unparalleled decisiveness. It collaborated with OpenAI, forged strong ties with other big AI companies, such as Mistral and G42, and above all invested enormously in data centres to develop the most gigantic models.

Thus, after a decade of stagnation, Microsoft's financial value increased enormously, to over USD 3 trillion in some periods. But unlike the era of the de facto monopoly of Windows and Office, this time Microsoft's position was less protected. All it took was for a small Chinese company to show itself capable of doing what OpenAI did with much less investment to send shivers down the market. Not least because DeepSeek was only one of the threats. A question mark hovered over the market and became more and more pressing. Despite investments amounting to hundreds of billions by the big American IT companies, generative artificial intelligence was still in many ways a solution in search of a problem. The game was therefore twofold: the best products had to be made, but above all, their adoption had to be encouraged. In this respect, the paths of Microsoft, Google, Amazon and others were different. Who would find the right one?

Probably, the answer will not be found in computing itself. The complexity of the world that has developed around the Internet, revitalised by artificial intelligence, is not comparable to the context that Microsoft has experienced in the half-century of its history. Geopolitical balances, made unstable by the most self-referential autocracies and by democracies that sometimes imitate their behaviour, come into play. Problems of incompatibility between systems of rules come forward. Corporate strategies are redefined. The need to reformulate the systems by which societies educate their youth becomes clear. Innovation needs motivated not so much by industrial competition as by adaptation to new climatic conditions are addressed. The vision of who will be able to win in this context will not be that of those who seek to dominate the market and find a protected area from change. It will probably emerge from those who know how to fully appreciate the value of an era that calls for finding meaning in innovation. One fact is certain. These are interesting times.