Hot on the heels of killing its Windows Defender support for Windows 10, Microsoft is continuing its investment in AI software with Windows 11. Most recently, Microsoft has announced a new extended push into its integrated and dedicated Copilot AI support. Microsoft promises easier access and more powerful tools for the average user, but critics are wary of what adopting the technology might mean.
Building Their AI Dream
Microsoft is far from the only tech company to focus its attention on AI integration. It’s now become standard fare, with a huge influx of investment from businesses of all sizes in the last couple of years. With their most recent move, Copilot will become more broadly accessible, activating with “Hey, Copilot”, similar to existing functionality with Google and Apple AI.
Source: Pixabay
The new systems push seems targeted at agentic AI. This is AI designed to listen to a user’s command, after which it can open a piece of software and complete the task while the human focuses on other areas. In theory, this software could offer immense positive rewards for users, but as usual with AI, the promise and the reality are two very different things.
The Potential of Simplicity
The idea that agentic AI could be given a task and could complete that task quickly and accurately is too appealing for any business owner to ignore. This wouldn’t just save time for one employee; it could theoretically even make some employees obsolete, significantly saving money on wages. AI, however, has a poor track record for accuracy and reliability.
In large language models AI, the only way to use it properly is to carefully check every single source, and also understand why a source might not be of value. This requires much more work than just asking a question and having it answered, and similar limitations apply to agentic AI.
With Agentic AI creating a document like an Excel sheet, this output would have to be carefully checked by a human every single time the output is created. This human would need mastery of Excel; otherwise, the output couldn’t be trusted. In turn, this means every user leaning on agentic AI would actually need to have more expertise to reliably use the system, not less.
The idea of bigger being better is a common belief in some spheres of tech, and too often it’s more of a case of reinventing the wheel. Quite often, the best software keeps it simple because it understands its core function. Consider the entertainment sphere with platforms like casino online UK websites. Games on these websites, like Fire Tale and Big Bass Halloween, use evolutionary design to improve the important components that already exist. This is why these games are so widely played and so technically efficient: they’re improving what’s demonstrably important for users.
Agentic AI aims to add another layer of complexity to systems that very few users are equipped to handle. It’s not, as advertised, a simple platform to unleash that gets you exactly what you need. While it might be possible that it will achieve its stated goal in the future, this won’t be for many years yet, and until then, early adoption could prove more of a liability than a benefit.
