Summary

  • Microsoft is unsure of how to utilize Copilot and has been integrating it into various platforms without a clear strategy.
  • Microsoft Copilot is the only way to access free GPT-4, making it the best free large language model available.
  • Copilot's integration into Microsoft 365 shows potential for usefulness, but it remains to be seen if it will truly be a game-changer.

When it comes to AI tools, there are few as well-known as Microsoft Copilot. It's a branding that initially started under Github Copilot back in 2021 as an AI coding tool, though it later entered the mainstream when Bing Chat was rebranded. Now, Copilot is in lots of places and coming to more, but honestly, it feels like Microsoft doesn't know exactly what to do with it.

Microsoft Copilot is probably the best free large language model

It's the only free way to use GPT-4

No matter what, Microsoft has one thing going for it with Copilot that nobody else has: it's the only way to access free GPT-4. With ChatGPT, you get GPT-3.5 for free, and with Google Bard, it uses Gemini Pro, which is still pretty good. However, Microsoft has very much shot first and asked questions later, just like Google did when it came to Google Bard as well. Companies saw how ChatGPT became the fastest-growing application in the world and wanted a slice of the pie with their own LLM technology.

Microsoft is slightly different, though, as the company reportedly owns 49% of OpenAI and leverages the power of its ChatGPT technology to power Copilot. With that said, Microsoft disputes any ownership stake in OpenAI, telling us that "While details of the agreement between Microsoft and OpenAI remain confidential, it is important to note that Microsoft does not own any portion of OpenAI and is simply entitled to share of profit distributions." Nevertheless, GPT-4 is probably still the best consumer-available LLM, and the company also sees it as a major boon to getting ahead in search for the first time. However, that's not enough for Microsoft. The company has since put Copilot into Windows, with rumors suggesting that Windows 12 will be built around AI. On top of that, Copilot is now in Microsoft first-party Office tools as well, so long as you have a Copilot Pro subscription, that is.

Regardless, Microsoft has been shoving Copilot into everything, and arguably, a little bit too quickly for fear of overwhelming consumers. I already suspect that Microsoft shifted away from the Bing branding because of the fatigue consumers faced in repeatedly seeing the brand in negative contexts following its launch, where it sometimes attacked and insulted users. Also, Bing has always been considered second-class to Google, and that's a hard reputation to shake.

Despite the technology, Microsoft doesn't know what it's doing

Launching as Bing Chat proves it

To be honest, I'm still surprised to this day that Microsoft decided to launch one of its most important technologies as "Bing Chat." Bing has always been seen as a second-class citizen to Google, being the butt of countless jokes over the years. For most people, Bing is the search engine they use when they boot up a Windows machine for the first time, open Edge, and type "Google Chrome" in the URL bar, where it then searches for a competing browser on Bing.

With that launch, Microsoft shot itself in the foot from day one, and the rebrand to Copilot makes it clear that Microsoft needed a change. However, even then, the company still isn't sure what it wants to do with Copilot. It's being integrated into Windows as a chatbot in Microsoft 365 and a web UI; however, powerful AI isn't the same as useful AI. Why would I ask Copilot to launch an application on my PC? Cortana could already do that, and we all know how that turned out.

Microsoft wants to lead the pack in artificial intelligence, but Copilot isn't likely to be the way to do it. It's cool and useful in many contexts, but I suspect that the Copilot button that will be available on keyboards won't always be popular. The Start menu button stuck around, but Copilot's own button gives the same vibes as Toshiba did when it tried adding a Cortana button to its laptops. At the very least, it's replacing the Menu key on keyboards... and who uses that key?

I know I keep beating the drum of Cortana, but to most people, Copilot in Windows will be the exact same. It's another menu where you can speak aloud or type to a bot that will either help you do things on your computer or forward you to a search engine. There's not much of a difference to the end user.

But Copilot is coming to Microsoft 365, too

Even then, it feels hamfisted

If you're wondering about Copilot launching in other applications, too, like Microsoft 365, it's hard to call that a calculated decision either. Microsoft is simply throwing Copilot at everything where it could potentially stick and help users out. To be clear, there's nothing wrong with doing that, and companies have done far worse than throwing random, powerful features at applications. Having said that, for Microsoft to give a cohesive user experience with Copilot, it needs to be a cohesive and well-thought-out product, too.

When it comes to using Copilot in Microsoft 365, it'll be able to help users do anything, and that's something I'm looking forward to. With PowerPoint, you can supply a Word file and ask it to make a presentation based on the document, and it'll do that, though that's a niche feature that won't be for everyone. It's infinitely more useful than anything that will end up integrated into Windows, though, and that's the point. Nearly all AI use cases are niche use cases, as is the nature of AI.

Copilot in Microsoft 365 is only starting to roll out, so it remains to be seen if that is more useful than simply asking Bing Chat to do X or Y and pasting the result into a program. Regardless, Microsoft is pushing Copilot wherever it can as fast as possible rather than focusing on individual features that can be powered by Copilot instead.

Pushing Copilot to the limit

With all that said, I think that Microsoft wants to push Copilot into as many places as possible, but that doesn't necessarily mean it knows what it's doing or has a specific game plan. It feels like Copilot will be something big somewhere, but Microsoft doesn't even know where yet and wants to try to find that out. I'm the most optimistic about it in 365 more than anything else, but I still suspect that we're a long way away from Copilot being the game-changing experience it feels like Microsoft wants it to be.