The introduction of ChatGPT is a game-changer, shaking up the software landscape in a big way. It's influencing a lot of what we traditionally understand as software and how it caters to users.

Take Copilot in Microsoft 365, for example. It's making employees more effective by giving them relevant info in emails, videos, chats, and even helping whip up documents or presentations. And let's not overlook its knack for answering both general and company-specific questions from employees.

Now, the cool part is that OpenAI and Microsoft are letting us build our very own customized ChatGPTs and Copilots. How might this shake up software development and, in turn, impact the end user?

Builders, the new revolution

We're seeing Microsoft introducing Copilot into all its services. Initially pitched as an assistant to boost employee efficiency, it's now not just one Copilot but several, each tailored to specific services in Microsoft 365 and more.

In one of my recent articles titled Am I Alone in Wondering How Many Faces Microsoft Copilot has? (you can read it here), I explore whether Copilot is a single unique thing with many faces or a bunch of unrelated services.

OpenAI has a cool feature that lets anyone create their own GPT. Think about it—your personal legal helper, language guide, or even the rule-keeper in a game.

Moving ahead, Microsoft introduces Copilot Studio, a tool that lets you make your own Copilot while keeping everything within the company. Imagine the ways this can help employees fine-tune Copilot for specific and concrete tasks.

Would be "discover" the new soft skill?

In the coming year, employees will create their own Copilots to fit their needs. During this busy time of creating, some Copilots may not work well due to setup issues, lack of usefulness, or not enough promotion within the company.

We also need to consider how people will use these Copilots—what problems they'll solve and what answers they'll get. It's crucial to have clear goals for customized Copilots. Are they for answering questions, creating documents, making audio or video, or working with tools like Word or PowerPoint? This plan might change over time, especially if the Copilot engine improves on its own.

This change might shift applications away from clicking through menus to more natural talking. Users could say things like "upload a document", "add info to a document with ID 2", or "filter by area and division". All by talking, not clicking.

In the future, everyone should be able to discover what commands Copilot understands and can do. From my experience as a user and creator, about 20% of an app's features get used a lot, compared to the almost unused 80%. Copilot might increase that number because it's easy to talk to and discover new commands just by asking what it can do or if it can help with something.

Wrapping up

Generative artificial intelligence is changing the way we work. ChatGPT has made it accessible to everyone, and Copilot has brought it into the business world. The rise of GPT builders and Copilots is a small revolution, shifting GenAI from general use to specialization.

The impact on us as users is still uncertain. However, communicating in natural language might unlock opportunities that were overlooked before due to lack of interest or complexity. In the coming days, we'll find out if the ability to "discover" is a valuable skill that gives some people an edge in using GenAI over others.

What are your thoughts on these developments? How do you envision the role of GenAI in the future of work? Share your opinions or any insights you might have on this evolving landscape.

References

• Introducing GPTs: https://openai.com/blog/introducing-gpts
• Microsoft Copilot Studio: https://www.microsoft.com/en-us/microsoft-copilot/microsoft-copilot-studio