The Blueprint - Part I: Leading With Knowledge & Culture

Taking an agile approach to implementing generative AI

clean pixel art of various data models, neon colors --ar 16:9

Hello and welcome to part one of my Blueprint series where I’ll be exploring where to start when building a strategy for implementing generative AI in business.

In this article I’ll cover the challenges of building a strategy around generative AI but also leave you with some practical tips to start thinking about.

Generative AI is still super shiny and new. It’s buzzy, it’s evolving at a rapid pace and it can feel like a full time job just keeping up with it all. What is clear to most people by now is that generative AI is an important, transformative and disruptive technology that is here to stay. It’s going to have a huge impact on society, culture and most industries. Business leaders, no matter the size of their company or the industry they operate in, should be thinking about their AI strategy right now.

In this first instalment in The Blueprint series I’ve taken inspiration from a recent Harvard Business Review (HBR) article - “Build a Winning AI Strategy for your Business”¹. It’s a highly recommended read, and I completely agree with the sentiment shared by the author:

“In my entire career in tech, I’ve never been more excited and optimistic than I am now.” - Christopher Young

Now that AI is moving from its auto-pilot phase to its co-pilot phase it’s more important than ever for businesses to start embracing and experimenting with this exciting new technology. However, this won’t be enough - to take full advantage of generative AI business leaders need to start setting out a strategic and agile roadmap for how they can integrate it across all the different parts of their business. Every roadmap will be different, and every business will have different types/levels of resources at their disposal, but my ambition for The Blueprint is to lay out the steps that any business can follow when thinking about their generative AI strategy.

The Challenge of Generative AI

The challenge of building a strategy around generative AI for businesses right now is that it’s still a nascent technology with many unknowns. As Young states in his HBR article, generative AI is “…the next wave of truly transformative technology with potential we cannot yet fully envision or appreciate.”

This makes it incredibly difficult to take a typical product/user-led approach. It’s hard, if not impossible, to get user feedback and envisage the breadth of uses for such a transformative technology. So where do we start? I have two suggestions:

1. Culture

Put simply, we don’t know what we don’t know. As such, an experimental and agile approach needs to be taken with generative AI. This will require businesses to have an open culture of innovation, iteration and continual learning.

The people in the day-to-day of a business, where the benefits of generative AI will most likely be felt, should be empowered to experiment with clear guidance and guardrails. I recommend businesses have three simple guidelines in place for the use of generative AI:

1. Don’t input anything confidential nor with copyright into a generative AI platform.

2. Don’t use generative AI for anything outwardly facing without approval.

3. Do experiment!

Empowering day-to-day teams is my recommended approach for a business to accelerate organisational knowledge and learning around generative AI. Encourage your teams to try out the various platforms as part of their regular tasks, get them to try different ways of asking questions, and point them towards some of the Guides & Learning on The Blueprint, especially the prompting resources.

Taking this approach will also instil optimism for generative AI across the business. Research from Boston Consulting Group² (BCG) suggests that familiarity with generative AI tools fosters optimism, with regular users being nearly twice as optimistic as non-users (62% vs. 36% respectively).

To facilitate this culture, business leaders and technology teams need to have open and regular dialogue with the day-to-day teams that are experimenting. Invite ongoing user feedback and put processes in place to capture learnings. These learnings can then be fed into an agile and evolving business strategy for generative AI.

This agile approach will be much easier to adopt for smaller businesses, but for larger global organisations that usually roll-out new technologies top-down or centre-out this will represent an uncomfortable departure from their tried and tested approach. These are exactly the findings of recent research by Microsoft³ that highlights a gap between what leadership and technology departments consider to be priorities and the actual needs of the workforce. Microsoft found that 61% of employees they surveyed are not an integral part of digital transformation projects and 70% say that organisational policies limit their ability to proactively explore or implement digital solutions on their own.

Businesses will need to tackle these cultural hurdles to fully harness the power of generative AI. The most successful ones will have their teams experimenting with generative AI to identify what they don’t know and in turn feeding their experiences back up to leadership and technology teams to then implement as part of a wider transformation agenda.

2. Knowledge

One undeniable truth about generative AI is that it is fuelled by data, and large quantities of it! So for businesses to get the most out of generative AI they need to find a way to allow generative AI to get the most out of their data. In fact, proprietary data is one of the two areas that Ilya Sutskever, the Chief Scientist at OpenAI, said developers should focus on when building generative AI applications in a recent interview⁴.

However, I’d argue that when it comes to generative AI, we need to shift our perspective on data and start thinking of it as knowledge. When you think about building a knowledge-base, as opposed to a database, to train generative AI it opens up your thinking beyond the obvious datasets you might typically use. For example, in most knowledge-based businesses, this ‘knowledge’ is currently locked in emails, Slack channels, documents, spreadsheets, presentations, databases, finance and people systems - a whole variety of different places.

All of these knowledge stores are incredibly valuable and generative AI can make them more accessible, and in turn create a huge amount of value from them. However, this wide variety of data does present a few challenges that will need to be worked through:

• How do you make all this knowledge available to a generative AI models?

• How do you keep a model up-to-date with knowledge that is fluid and fast-moving?

• How do you maintain security and confidentiality when opening up these knowledge stores to a generative AI model?

• How can you organise, harmonise and label all this knowledge so that it can be used?

These are all important questions that businesses will need to address in order to fully embrace generative AI. We’ll be addressing these questions and more in future instalments of The Blueprint.

The other side of knowledge is employee education. There is currently a big skills gap in organisations and according to the research from BCG, 86% of employees believe they will need upskilling but only 14% have received any formal training. This is a huge gap to close, but most businesses don’t know where to start. There are plenty of questions to answer, such as:

• How much background knowledge do employees need on generative AI?

• How do we support our teams to start using generative AI?

• Should we be training technical teams or business users, or both?

• Do employees need different training for different generative AI platforms?

One of my main motivations for launching The Blueprint is to start upskilling leaders and teams in businesses and to help them better understand generative AI. Alongside this series I’ll also be posting educational content in The Classroom that will give readers the foundational knowledge they need as well as practical tips on using generative AI platforms.

The Potential of Generative AI

Start imagining the opportunities that could arise from having generative AI models trained on your businesses’ entire knowledge-base and how new tools based on these models would be embraced by a workforce that is actively engaged in the transformation process. It’s an exciting and enticing thought!

By getting the culture around transformation right and thinking about all the knowledge across the organisation, business leaders can lay the foundations for the successful and fast adoption of generative AI. I believe it’s vital for businesses to get this right as those with the best generative AI strategy will benefit from competitive advantage and also attract the best talent in their industries.

In this article we’ve covered some important, foundational topics to help business leaders put together the start of their strategy for generative AI:

• Start by getting the culture right across the business - put the right guidance in place and encourage and empower your day-to-day teams to experiment.

• Invite ongoing user feedback and put processes in place to capture learnings.

• Have an agile and iterative strategy for generative AI that can evolve and adapt over time.

• Start thinking of knowledge, not data, and consider all the different stores of knowledge across the business.

In the next article in The Blueprint series, we’ll start answering the questions outlined above around managing different knowledge stores across the business and making them available to generative AI models. We’ll discuss the hurdles businesses face when dealing with dynamic data, preserving privacy, and ensuring security. We’ll also explore what these factors mean for training generative AI models.

---

“The future is already here, it’s just not evenly distributed.“

William Gibson

---

This article was researched and written with help from ChatGPT, but was lovingly reviewed, edited and fine-tuned by a human.

---

Found this article helpful? Share it with your network to help them navigate the challenges and opportunities of implementing generative AI.

1. https://hbr.org/2023/07/build-a-winning-ai-strategy-for-your-business

2. https://www.bcg.com/publications/2023/what-people-are-saying-about-ai-at-work

3. https://www.microsoft.com/en-us/worklab/four-ways-leaders-can-empower-people-for-how-work-gets-done

4. https://youtu.be/xym5f0XYlSc