Introduction

Like many new technologies before, Artificial Intelligence (AI) provokes strong reactions.Some are eager to try every life-changing hack, while others fear AI will replace them—or worse—take over the world. Bad sci-fi plots aside, love it or hate it, AI is here to stay. And it’s causing dramatic changes in how we train and learn at work.

AI—specifically generative AI models like ChatGPT, Midjourney, and Bard—have the potential to unlock new possibilities for learning and development pros. Think of these models not as a replacement but as an enhancement--even a superpower. And just as Tony Stark tested prototypes before he truly unleashed Iron Man, now’s the time to learn, test, and enlist superpowered learning.

Your organization is already seeking answers, so it’s time to lead it into the future.

This e-book examines best practices for adopting generative AI into learning and development roles—and how AI can actually strengthen your value to the organization.

Generative AI at a Glance

For over a century, machine learning was restricted to labs and used to predict and classify patterns. That changed in the 1970s, when Dr. Geoffrey Hinton, often called the "godfather of modern AI," proposed a revolutionary idea. He believed unlocking AI’s potential meant leaving behind the logic-based, symbolic systems his contemporaries favored and instead building networks that mimicked the human brain. Think of his proposal as moving from pulling information piecemeal from a filing cabinet to casting a wide net to select the correct information from all the file cabinets simultaneously. This network of information, patterns, and interactions—like the brain uses—would make AI faster and more comprehensive. Hinton’s theory led to "neural nets" and the generative AI in headlines today. This type of machine learning can predict and classify patterns like before, but it can also draw on that information to create something new.

It makes large language models (LLMs) like ChatGPT and Bard and text-to-image or text-to-video systems like DALL·E 2 and ElevenLabs possible.

These publicly available, generative AI models are typically "trained" on content from the internet. Users ask or "prompt" the AI with questions, and AI gives unique responses based on what it’s learned.As you can imagine, that training has pros and cons, which we’ll discuss in this guide.

In addition to public models, organizations can train generative AI on information specific to their organization. Some models are embedded within existing tools to seamlessly assist employees in the flow of work. For example, a payroll company might train an LLM on their unique payroll systems and embed the trained LLM within their system, making it easy for employees to ask or "prompt" the LLM for help and get an immediate response relevant to the organization.

system, making it easy for employees to ask or "prompt" the LLM for help and get an immediate response relevant to the organization.

Do You Use AI at Work?
_{(Select your response to view results)}

No, and I don't want to
No, but I plan to try soon
No, AI use isn't allowed
Yes, I use AI occasionally
Yes, I use AI frequently

AI's Outsize Impact

If you're using AI, you're not alone

AI adoption is up 2.5x
over the past 5 years¹

Estimated $4.4 trillion value to the
global economy annually²

80 percent of U.S. workers
may be impacted by LLMs³

Sources:
McKinsey¹McKinsey² Brookings³