Understanding Gen AI in Business and Strategy

Is generative AI good or bad? 

Many in the artificial intelligence field constantly debate this question, but is there be a better way to approach it?

The short answer is “Yes!” We should be figuring out the right way to implement AI tech into our strategy and processes.

Lisa Esch, Senior Vice President and Chief of Strategy, Innovation, and Industry Solutions at NTT DATA provides a crash course on all things generative AI, LLMs and the risks and calculations any type of organization should be factoring into their decision-making surrounding artificial intelligence. 

Her experiences in healthcare tech — and the lessons she’s learned from witnessing how AI has shifted the industry’s landscape — provide an excellent case study for anyone asking how this game-changing technology can fit into their operations.

The landscape of gen AI: what exactly is it?

There tends to be a misconception that generative AI is newer than it is. Gen AI was not born in January — nor when the general public first started hearing about it — but has been developing for quite some time with larger corporations investing in it. 

But what exactly is gen AI?

“Generative AI is this process of taking information, putting it together with data from other sources, and really focusing on creating answers. The key is generating new things — new ideas, new pictures, new thoughts, new concepts and so on,” says Lisa. 

Today, generative AI can be summarized into three main categories: conversational AI, foundational models, and large language models (LLMs).

Diving into the 3 main categories of AI

  1. Conversational AI: Many people are already familiar with conversational AI, or chatbots. There is a strong market for chatbots—especially in customer service—as they can help remove human barriers to accessing information and getting answers quickly. 
  1. Foundational Models: This type of AI goes beyond what conversational AI can achieve by taking data sets and information and training itself to create new output. Companies like Google and Microsoft have been using foundational models for training, learning, and creating content because these models create more than just language content. They’re broad and can create pictures, imaging, etc. in addition to written content. 
  1. Large Language Models: LLMs are where a lot of major organizations are focusing their time, energy, and money right now. These are AI models trained on text—words and information—and focused on creating human-like interactions through text. 

Buy vs. build: which is better when it comes to AI strategy and implementation?

If your company hasn’t started its AI journey yet, you may be wondering if it’s too late, if you’re already too far behind. 

The good news is it’s not too late—you might actually be starting at a better point now than you would have earlier!

When Gen AI first started, it seemed like everyone went out and bought a bot. What ended up happening for a lot of them was that they took inefficient processes and automated them. 

Ultimately, inefficiently automating things at 10 times the speed wound up causing more problems than solutions.

Companies then started recognizing the importance of first looking at the process being automated and fixing it if it needed to be fixed before applying the automation to it. 

You don’t want to be automating things that are still inefficient, but the good news is now there are plenty of companies that have experienced this process and can provide some guidance for others. 

The answer to the build vs buy questions becomes clear when you consider how you can take established models and train them to make them safe for your organization and able to do what you need them to do with the information given at a highly successful rate.

However, it’s not quite as simple as picking a model and running with it. 

Listen to The Strategy Gap

A podcast about the space between savvy strategy and practical execution, including everything that can go wrong on the way. 

The importance of partnership, alignment, and prioritization in AI strategy building 

“When thinking about where to get started and what to do with gen AI, it’s important that we really are thoughtful about how we use it because there are some risks associated unless you’ve got a closed foundational model,” says Lisa.

She encourages those starting to develop their AI strategy to find a foundational model and partner to provide necessary guidance and mitigate risks. 

“Understanding the complexity of gen AI is way out of most organizations’ wheelhouses. Partner with organizations that have deep experience with security, compliance, law, ethics and the bias that happens in AI models, someone that can help you build a strategy and move around it,” Lisa says. 

Many organizations implementing generative AI today are doing so to fix a specific problem but lack the strategy needed to effectively reach their goals. 

So, how should they go about building a strategy for something so new and ever-changing?

3 steps to building an effective AI strategy

  1. Establish governance: First, you have ethics and compliance defined and have new people on your team constantly engaged with what you’re doing in the AI space. A high level of involvement will help with the element of human interaction and dealing with legal, compliance, etc. People are going to need to be more involved because there’s a specific problem to solve.
  2. Determine the problem you’re trying to solve: Consider what solving the problem looks like. What is the ROI, and how are you going to measure that? Pick your strategy and foundational model partners as part of this discussion to ensure alignment. Then figure out how to set it up, manage it, govern it and keep it, your organization, and your customers safe before you build out your prototype.
  3. Take a step back: Stop and take a moment to think about where gen AI is going to have the biggest organizational impact. Think about how you can use this technology in a way that makes your business more efficient, better, stronger and helps you grow. 

Moving from a point solution to considering AI’s long-term impact

When it comes to gen AI strategy, there are two paths: the big transformational play where AI can take care of it all and the single-point solution where you have significant pain points you need to solve and you can solve them immediately. 

How should you decide which path to take?

“Take a deep look at your business. What parts are spent on just processing, like coding, or other basic transactional types of things? Those are the things that you can really tackle with gen AI to transform your business and really automate your process,” says Lisa. 

It’s important not only what you do with your strategy roadmap but also what you don’t do.

Consider the impact, outcome, financial ROI and culture or people change that might be required to take each path before deciding if a point solution will be enough or if more might be necessary later on.

To hear this interview and many more like it, subscribe on Apple Podcasts, Spotify, or our website or search for Generation AI in your favorite podcast player.

👋

Meet the Author  Jonathan Morgan

Jonathan Morgan is the VP of Revenue Operations and Head of Marketing at AchieveIt. Jonathan has spent time in roles across strategy consulting, sales, customer engagement, marketing, and operations, enabling a full picture view of strategy & strategy execution. His generalist background encourages a full picture view of strategic planning & strategy execution. Jonathan graduated from Georgia Tech and received his MBA from the University of Florida.

Hear directly from our awesome customers

See first-hand why the world's best leaders use AchieveIt

See AchieveIt in action 

Stay in the know. Join our community of subscribers.

Subscribe for plan execution content sent directly to your inbox.