fbpx
Andrew Murphy

Andrew Murphy

I'm Murph, Head of Strategy at jaam. My goal is to help companies, departments, and individual users be more effective through automation by combining great people and great technology.

More blogs and news from jaam...

Microsoft Copilot versus ChatGPT

Navigating the Synergy of Microsoft Copilot and ChatGPT: Insights from the jaam automation Hackathon

AI hackathon

Images above and below all created using DALL-E on ChatGPT

At jaam, we use artificial intelligence (AI) every day, both in our customer automation implementations and in our own internal operations. The landscape of AI is constantly evolving, with new tools and technologies being released frequently. As automation experts, the jaam team are well versed in what is available and how it can be used and, at the moment, generative AI is a hot topic.

Our CTO, Graham, recently penned a blog on exactly what generative AI is and why it’s important. Out in the field, our clients and contacts are keen to find out more about generative AI, and one of the questions I am frequently being asked is: “When should you use Microsoft Copilot versus ChatGPT?”

The answer, though seemingly complex, is quite straightforward when we understand the distinct advantages each platform offers and the business needs you are looking to apply an AI solution to.

Strengths: Microsoft Copilot vs ChatGPT

AI superhero

Microsoft Copilot, seamlessly integrated into the Microsoft technologies, including the Office suite, GitHub, and the Power Platform, offers a unique advantage for users entrenched in Microsoft’s ecosystem. Its ability to provide context-aware assistance directly within these tools makes it a great ally for specific tasks. For instance, if you’re venturing into creating your first Power Automate flow, or drafting a PowerPoint presentation, Copilot’s integration offers unparalleled convenience and efficiency.

Conversely, ChatGPT‘s prowess lies in its ability to digest, analyse, and synthesise large volumes of text-based information. When faced with the task of summarising extensive documents, drafting proposals or getting help with coding, ChatGPT stands out. Its capability to process and generate content from vast datasets makes it a powerful tool for comprehensive analytical tasks.

Synergy: Harnessing Microsoft Copilot and ChatGPT together

AIs hugging

The distinction, however, doesn’t imply a binary choice between the two, but rather suggests a complementary relationship.

This synergy was vividly showcased at a recent hackathon hosted by jaam automation for a prominent tech company. Aimed at demystifying AI and honing in on generative AI, the event saw 10 teams, each consisting of 6-8 members, tackle challenges that necessitated a blend of Copilot and ChatGPT’s functionalities.

The challenges, designed to cover three core areas—customer research, data set analysis, and proposal generation—underscored the collaborative potency of Copilot and ChatGPT. Moreover, to inject a dose of creativity and team spirit, participants were also tasked with devising unique team names and mascots, a testament to the versatility of generative AI in fostering both analytical help and creative expression.

During the hackathon, teams discovered that while Copilot excels in crafting PowerPoint presentations and performing basic document analysis and email drafting, ChatGPT proved superior for extensive document and data set scrutiny. A notable challenge encountered was Copilot’s lack of conversational context continuity across different Office products, which resulted in a somewhat fragmented experience. It’s an area ripe for improvement, and we’re hopeful Microsoft will address it.

On the fun side of team mascots (apart from teams running out of tokens 😊) when it came to image generation, some teams found alternative AI engines, such as Midjourney and Stable Diffusion, delivered better results with fewer spelling errors compared to DALL·E, the ChatGPT image creation tool.

The overarching lesson from the hackathon was clear: in the current generative AI landscape, a hybrid approach employing multiple tools is essential for maximising efficiency and effectiveness.

One final word of caution, data integrity and privacy is critical for businesses, so please make sure you are aware of whether you can pass content into the domain of public generative AI models, but that topic’s another whole blog on its own!

To learn more about the uses for ChatGPT and Microsoft Copilot in a business context, speak to the experts at jaam automation. Get in touch.