Betting the Company on Uncertain Returns
Microsoft (and several other companies) push AI technologies hardThrowing sums of money and unprecedented work in the apparently bottomless mouth of the AI. It is not an exaggeration to say that it seems that the company bet everything on AI.
However, with the revelation that the current AI technology will not develop quickly enough to give a return to these investments – and perhaps never – and that the vast majority of AI pilot programs in the real world have so far failed, is the venerable old software giant about to innovate with business?
Microsoft’s all-in bet on AI
Microsoft injects AI in all corners and corners that it can find in its existing software and services. There is a dedicated co -pilot button on surface devices, there is the AI in each subscription, the co -pilot is cooked in the heart of Windows with (controversial features such as the recall), and it is in each application grouped of services like Microsoft 365.
Some of these tools are, of course, really useful and will make you more productive during the presentation, but at this stage of the game, it could also likely hallucinate errors in a spreadsheet.
However, the usefulness of AI for end users like you and I may not be the most salient point here. In 2025 only, according to the Microsoft blog, the company should invest $ 80 billion in AI. The AI is always incredibly at high energy intensity and is not particularly profitable. Local AI treatment is limited, and companies like OpenAi always lose money, even with expensive subscriptions. I have seen speculations according to which the massive Xbox launch are at least partly motivated to release investment resources in AI.
However, we have studies like a recent MIT which revealed that 95% of companies leading AI pilots have not seen any return on investment. Of course, the study has its criticisms, but there is no prejudice that the reality of AI and the beaten media and the expectations for this are not yet tail. Sam Altman, CEO of Openai, admitted that AI could be in a bubble, which, in your opinion, would give Microsoft a break.
But, with regard to anyone outside can say, it seems that Microsoft is locked up on the AI - for better or for worse.
When integration begins to cannibalize
For me, it seems that there is a fairly large intake of self-deficiency if the AI is infused in each part of the Microsoft product menu. Over time, has “assistant” in Excel or Outlook, or wherever it ends, will simply start with jobs from the software with which it is grouped. Why bother with Excel when Copilot could potentially ingest your data and throw the spreadsheet you want?
It may be the long-term plan, but ultimately with a more advanced form of co-pilot, it could actually be the only product and service that Microsoft sells. There is no application ecosystem, just an AI frontal that takes requests and spits what was used by individual applications. Maybe it seems a good thing to you, but for me, it also looks like a business that hurries to dismantle a robust ecosystem without anyone asking for it.
Ecosystem lock in relation to the erosion of the ecosystem
Microsoft has built its empire on the locking of the ecosystem. MS-DOS and Windows did not become dominant in the office space because they were technically the best. It is largely the result of Microsoft which intelligently fits into the right place and at the right time, which leads to a chicken and eye situation where everyone uses their operating system, which means that the software is mainly produced for this operating system, which means that more people must use a Microsoft operating system to execute the software they need.
An AI service based on the cloud is not really the same with regard to the locking of the ecosystem, isn’t it? If something like Copilot becomes the main product and service, that’s another reason why you don’t need to be a Windows user. You can just use Copilot from the Internet interface you want. Likewise, it would be much easier to jump between different AI services, and it is not clear exactly how Microsoft would force people to stick to the generator AI instead of going with the brightest flavor of the month.
The risk of reaction of users and developers
Lucas Gouveia / Geek. Microsoft | brain_fix / shutterstock
Even if Microsoft has planned everything and knows something about the medium -term future of the AI that we do not have, there is always the problem of reaction of their customers and people who develop software to operate on Microsoft operating systems. We have already seen lots of offices with regard to the co -pilot and how people just don’t want them on their computers.
Microsoft had to find this surprising, but he returned to the drawing board in order to obtain features such as the recall in a form that people will accept. What Microsoft is not done is to let it up. He will continue to try to make the piece of the Puzzle Ai in shape, and if you look at what extent it is sunk in AI, I suppose it makes sense.
What about developers who still create software applications for Windows users? Something like Copilot and all that is under this AI umbrella could be the ultimate form of sherlocking. Why develop individual applications when the AI that lives at the heart of your operating system could potentially perform an arbitrary function you ask for it?
What it takes to Microsoft to avoid breaking
The solution is not to slow down the adoption of the AI - it must position it properly. Copilot must be an amplifier, not a replacement, which facilitates workflows without removing the applications on which users depend.
Gardement is not only to prevent AI chatbots from saying naughty or offensive things. Microsoft must define the limits of its IA implementation clearly and early in the game, so that no one wonders if it is finally time to go to Linux instead.
