Regarding Microsoft's position on AI, its CEO Satya Nadella has made it clear: they like it. They like it a lot. In the company's annual report, in a letter to shareholders he praises AI in every way. And it's not hard to see why: He clearly believes that this is the biggest, and perhaps the only, advance in computing that he has really mattered in more than a decade.
Although Microsoft's business is strong, you might think they've been hanging around. Its bids to enter the mobile, search and hardware sectors have stalled or failed, and numerous product experiments have failed to penetrate their respective markets.
Therefore, it is normal that for the CEO, iA is the only advance in the market since in the previous ones, until now, all of his competitors have triumphed.
On the other hand, the cloud business is solid and the company and its products have been increasingly shaped around that premise, just as everyone has done due to market demand. But even that success was beginning to weaken, because no matter how profitable it is, the space for innovation is limited.
Without a doubt, for years they must have been attentive to trends, observing whether one novelty or another was worth adopting. Social networks? No, too much work. Physical aptitude? The infrastructure and technical user elements do not support it. Blockchain? Redundant and risky. Metaverse? Very funny.
Microsoft after entering and failing in all of them, thanks to its inherited captive market, had to wait. The AI wave appeared and they started rowing for their lives.
Exact place, exact time
As Nadella writes in his annual letter:
This next generation of AI will reshape all software categories and all businesses, including ours. Forty-eight years after its founding, Microsoft remains an important company because time and time again (from PC/Server to Web/Internet and Cloud/Mobile) we have adapted to technological paradigm shifts. Today we do it once again, as we lead this new era.
This is followed by a couple dozen examples of where AI is being implemented across its business units, products, and long-term efforts. This is not a hobby for Microsoft; In reality, they have decided that this is the next phase of personal and business computing.
And it's not just an enabler, either, like a silicon breakthrough that makes data centers run twice as efficiently, or a battery that lasts twice as long. It is, so to speak, a ecosystem:
The long arc of computing has been shaped, in many ways, by the search for increasingly intuitive human-computer interfaces: keyboards, mice, touch screens. We believe we have reached the next big step forward: natural language, and we will quickly go beyond it to see, hear, interpret and make sense of our intent and the world around us.
You can almost see the stars in their eyes: you have to imagine being at the helm of a major technology company like Microsoft during a transformation of this magnitude. They have already dabbled in the idea of leaving the mouse and keyboard behind, but so far their natural language interfaces (like Cortana) and alternative hardware (like HoloLens) have not risen to the level of parlor tricks as discussed above.
But, whether by good luck or foresight, they backed the emerging leader in Artificial Intelligence in natural language: OpenAI. Not only does the technology really seem like a real paradigm shift, but, for example, the way the cookie It put them in a suitable situation to give their eternal rival Google a black eye. Google itself, the most preferred search engine, for its part, has been surprised by the rapid shift towards AI, despite having internally created the concepts that allowed it. They're trying to bounce back, but the company has always struggled to successfully unite behind a unifying concept, and this time may be no different.
This alliance between Microsoft and OpenAI is liberating for both. OpenAI gets a combination of investor and customer with effectively bottomless pockets and a sincere desire to integrate AI tools into every corner of their business. Microsoft is spared the embarrassing need to appear – as it actually is – far behind the curve in AI development, because it can simply present the market-leading product as its own. Nadella does not mention that Microsoft is training its own models, although they are probably doing so quietly to protect against betrayal, because their efforts pale in comparison to the momentum of their partnership with OpenAI.
We must try to place ourselves in the opposite situation: that it was Google who had made a fortuitous deal with OpenAI, leaving Microsoft out of the game. Microsoft would be even worse off than Google, since it would have to struggle to create LLMs even a fraction as good, and every month they spent trying to catch up, their competitor gained another million users.
Therefore, it should come as no surprise that Microsoft is spending huge amounts of money to strengthen its position and, where possible, expand and deepen its partnership with OpenAI.
Assess the risk intrinsic to euphoria
However, a disturbing note that Nadella highlighted was his characterization of the second of the two advances that he believes define this era of AI: “the emergence of a powerful new reasoning engine.”
If you're familiar with the way this generation of AI models work, you'll know that they don't use the reason, just like a calculator reasons when you ask him to multiply two numbers.
Presumably, Nadella is neither naïve nor uninformed on this matter. She knows what she is saying here: that these systems perform functions that in many ways are indistinguishable from reasoning. Asking a computer to summarize a long document and do it, or even do it in iambic pentameter, it seems magical, because until recently, only those with the ability to reason could do it.
It turns out that language patterns are predictable enough that some reasoning tasks can be reduced to statistical tasks. This is notable enough on its own that we need not overstate the concept.
But this language is indicative of the possibly undeserved trust that security systems Artificial Intelligence have created in sponsors like Microsoft. They are capable of a lot, but with only a couple of years of existence, they are still in their infancy. They will become more capable, yes, but we will also learn from their limitations, and possibly only when those limitations have already created serious harm.
As AI ethicists have repeatedly warned, the risks of AI are not a future apocalypse or theoretical systems that displace entire industries, but rather overconfident and misinformed applications of the systems we have now. An overzealous CEO can do a lot of damage with AI models that aren't inherently capable of doing it themselves.
The balancing act Microsoft must pull off is investing at a pace that puts them ahead of their competitors, but not so far that they end up in a minefield with everyone else watching from afar. The curse of the innovator (or in this case the integrator) is to be the first to confront new risks, and Microsoft seems ready to play this role by putting AI to work, as far as it is known, in almost every company. unit and product where it is conceivable to include it.
Where will you find a market? Where will it fail miserably? Where will it attract lawsuits? Where will its disappearance be regulated? Satya Nadella doesn't know, but he and his shareholders are going to find out one way or another. Things are getting exciting again.
