Is Microsoft Copilot just another AI tool, or a major leg-up for current Microsoft 365 power users? Everywhere you turn, you hear people ask, “Is AI going to take my job?” So what’s the verdict?
The good news is that AI won’t take your job. But! Someone who knows how to use AI will. For example, 77% of developers who use GitHub Copilot say that the tool helps them spend less time searching for information, and 74% say it allows them to focus on work that matters to them.
Instead of labeling AI a rival, it’s time to make friends with it and see what it can do for you.
That being said, there seems to be a new AI tool launching every week, so it’s tricky to know which ones are worth exploring. If you’re a Microsoft user, you can’t afford to ignore Copilot.
What Does Microsoft Copilot Do?
Microsoft Copilot applies the power of large language models (LLMs) to the data in your Microsoft Graph, such as calendars, emails, chats, documents, and meetings. It will also be embedded directly in every Microsoft tool like PowerPoint, Excel, and Power BI (like Github Copilot) to help you quickly ramp your skills and deliver more sophisticated solutions.
The AI capabilities boost your productivity and unlock creativity by taking repetitive, mundane tasks off your plate.
Windows Copilot is built into Windows 11, allowing you to open a chat window directly on the desktop and ask the software to assist you with tasks or grab information. The tool is also embedded in various Microsoft 365 apps, such as Word, Excel, PowerPoint, Outlook, Teams, and more, to deliver a seamless experience while increasing productivity.
Bottom Line: Is Microsoft Copilot Useful?
Unlike ChatGPT, Copilot works with your datasets to generate outputs, reducing the chances that the program gives you an incorrect answer (or hallucinates). It’s important to note here, though, that since it’s working off of your existing data, your data infrastructure needs to be strong in order to get the best results. This is the core baseline of “Unified Data Analytics”.
You can also ask the program to help you improve how you use the data to produce more accurate insights.
Here are some examples of how Copilot works with popular M365 apps:
Outlook
Summarize long email chains, draft responses, flag important messages, and improve the tone or wording of your communication.
Teams
Set agendas, organize discussion points, create meeting notes, and formulate action points after a meeting.
Word
Create document drafts, rewrite paragraphs, summarize text, and improve the tone of your writing.
Excel
Analyze data, create reports, generate visualizations, and predict how changes in specific KPIs can impact business outcomes.
PowerPoint
Turn ideas into decks, visualize data, and automatically create speaker notes for a presentation.
Viva Sales
Scan customer relationship management (CRM) data and compare your sales collaterals with your competitors’ to optimize responses.
Power Platform
Build business solutions by defining an application with Power Apps in natural language or describing a workflow in plain English to implement it in Power Automate.
What are the Disadvantages of Microsoft Copilot?
Microsoft Copilot (and generative AI technology in general) are still in the early stages of development. While the tool can help you boost productivity and improve communication, you should be aware of a few caveats:
1. Inaccuracies and Hallucinations
If Copilot can’t find the appropriate information from your database, it may make up the answer to a question, and do it with extreme confidence.
Even though the technology is evolving and is more likely to tell you it doesn’t know the answer instead of fabricating one, you should be aware of potential hallucinations. To avoid this issue, ask for assistance to make your data better instead.
Additionally, it can’t account for errors in your data system. Since it’s working off of existing data, you’ll need to prep your systems to avoid any inaccuracies caused by incorrect data entry, duplication, etc.
2. Inability To Retain Conversations
Your conversation and the information gleaned from it disappear once you close a window. The software also deletes the existing data (i.e. a table) and restarts the conversation when you hit the refresh button.
Because of this, Copilot isn’t able to pick up a conversation where you left off after closing a window or learn from your interactions over time.
3. No Connection With Data Systems Outside Microsoft
Copilot can’t connect with existing systems that are not directly integrated with M365 or OneLake, which limits your ability to gain insights from all your information—creating data silos that may prevent you from gaining accurate insights and a holistic view of your data.
4. Security Concerns
Let’s say your organization uses Copilot to help build reports. But what if an employee requests information they shouldn’t have access to?
Copilot may not have all the controls required to restrict the data it pulls to ensure security and privacy compliance. There’s also the risk that the program misunderstands a question and pulls data from the wrong source.
5. Inconsistency in Communication
While Copilot can draw on branded materials, copy, and images to create new outputs, the content may not consistently represent your brand’s nuances.
The challenge multiplies if you rely on the AI program alone to deliver on-brand communications across various business units in a large organization. It’s always a good rule of thumb to use any AI tools as a helping hand, not a replacement that you leave to its own devices.
Preparing Your Data Infrastructure for AI
Insights from an AI tool are only as good as the data it has access to. While Copilot can be a helpful tool in building solutions to support your data processes, it’s not quite to the point where it can take the lead in analyzing your organization’s data. However, if your organization wants to be on the cutting edge and make the leap as soon as it’s released, here’s a great resource on how to prepare for Microsoft 365 Copilot.
As we incorporate more AI tools into the workplace, it’s more important than ever to have unified data to navigate AI implementation, and generate accurate insights across the organization.
So if you’re considering incorporating Copilot—or any other AI-powered tool, for that matter—into your organization, you first need to take a step back and make sure that you have the right foundation to support it. The tool is currently only available for private preview and will not be accessible to the public until later in 2023. So now is the perfect time to prepare your data infrastructure for the big shift.
