Exploring the Landscape of AI Productivity: Navigating Opportunities and Challenges

Written by Ian Morrish, Ian is our Microsoft ‘Guru’ leading the way for IL in all things Cloud and Infrastructure.

In the ever-evolving realm of AI, the spotlight is on the profound impact it could have on both personal productivity and organisational dynamics. In a recent discussion with our Microsoft Guru Ian Morrish, we delved into the potential implications, challenges, and strategic considerations.

Immediate Impact on Personal Productivity:

The immediate benefits for personal productivity are tangible, particularly in enhancing individual roles. Picture this - getting unstuck in Excel by seamlessly asking AI for the right formula.   Or using AI to prepare the first draft of a marketing brochure.

At IL, we anticipate that these personal productivity gains will be more achievable than gains delivered at an organisational level. This is due to the extra work required to implement appropriate content controls using Purview sensitivity labels.  Many organisations have not yet addressed the need to identify and manage sensitive information and so there are risk and governance implications to be considered.

Exploring Cost-Effective AI Solutions:

While there are out-of-the-box (OOTB) methods available from Microsoft, they come with additional licensing costs. There is a trend towards AI at the edge, where public models merge with private data. Copilot, as an example, represents this trend. Organisations need to consider whether the investment aligns with the productivity gains for them right now.

Internal Considerations for IM:

We’ve been actively brainstorming how we can leverage AI internally. The challenge lies in our relatively modest reference material pool. A conceptualised solution involves creating a database of 'overlapping chunks of reference material,' meticulously indexed for each document. This can then be referenced in the answers provided by a Large Language Model (LLM).

Short-Term Focus on Key Reference Locations:

By identifying key reference locations such as manuals, channels, and build documents, creating an AI-powered chatbot becomes possible.

Other immediate uses include enhancing personal productivity in applications like Excel, PowerPoint, and Word documents, especially in the context of translating documents to presentations.

Long-Term Outlook:

Considering the long-term perspective, AI at the edge presents an interesting opportunity. This approach envisions having control over AI, possibly hosted in Azure but with more cost-effective options, especially within the open-source AI movement. While Copilot remains a consideration, we are watching its adoption (licensing) rate and the potential for an end-user purchasable license in future.

Strategic Adoption and Uptake:

The current landscape of AI adoption is dominated by personal use cases. For organisational implementation, strategic questions arise, such as identifying design considerations and potential gaps in information that might hinder accurate AI responses.

Predicting Adoption and Mitigating Risks:

A critical consideration is the low adoption of Microsoft's Copilot compared to the widespread acceptance of OpenAI in Fortune 500 companies. Predicting the trajectory of adoption involves creating what-if scenarios, weighing quick versus slow uptake, and continuously reassessing strategies.

As we continue to navigate the dynamic landscape of AI, the key takeaway is the need for a nuanced approach. Balancing the immediate gains of personal productivity with long-term organisational benefits requires a strategic and adaptive mindset. The journey involves constant evaluation, learning from industry trends, and aligning AI investments with your unique needs and challenges.