These materials are © 2024 John Wiley & Sons, Inc.

Any dissemination, distribution, or unauthorized use is strictly prohibited.

 Enterprise
Generative AI
Writer Special Edition

by Alaura Weaver

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Enterprise Generative AI For Dummies®, Writer Special Edition

Published by
John Wiley & Sons, Inc.
111 River St.
Hoboken, NJ 07030-5774
www.wiley.com
Copyright © 2024 by John Wiley & Sons, Inc.

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any
form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise,
except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without
the prior written permission of the Publisher. Requests to the Publisher for permission should be
addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ
07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permissions.
Trademarks: Wiley, For Dummies, the Dummies Man logo, The Dummies Way, Dummies.com,
Making Everything Easier, and related trade dress are trademarks or registered trademarks of
John Wiley & Sons, Inc. and/or its affiliates in the United States and other countries, and may not
be used without written permission. Writer and the Writer logo are registered trademarks of
Writer. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc.,
is not associated with any product or vendor mentioned in this book.

LIMIT OF LIABILITY/DISCLAIMER OF WARRANTY: THE PUBLISHER AND THE AUTHOR MAKE NO

REPRESENTATIONS OR WARRANTIES WITH RESPECT TO THE ACCURACY OR COMPLETENESS OF
THE CONTENTS OF THIS WORK AND SPECIFICALLY DISCLAIM ALL WARRANTIES, INCLUDING
WITHOUT LIMITATION WARRANTIES OF FITNESS FOR A PARTICULAR PURPOSE. NO WARRANTY
MAY BE CREATED OR EXTENDED BY SALES OR PROMOTIONAL MATERIALS. THE ADVICE AND
STRATEGIES CONTAINED HEREIN MAY NOT BE SUITABLE FOR EVERY SITUATION. THIS WORK
IS SOLD WITH THE UNDERSTANDING THAT THE PUBLISHER IS NOT ENGAGED IN RENDERING
LEGAL, ACCOUNTING, OR OTHER PROFESSIONAL SERVICES. IF PROFESSIONAL ASSISTANCE IS
REQUIRED, THE SERVICES OF A COMPETENT PROFESSIONAL PERSON SHOULD BE SOUGHT.
NEITHER THE PUBLISHER NOR THE AUTHOR SHALL BE LIABLE FOR DAMAGES ARISING
HEREFROM. THE FACT THAT AN ORGANIZATION OR WEBSITE IS REFERRED TO IN THIS WORK
AS A CITATION AND/OR A POTENTIAL SOURCE OF FURTHER INFORMATION DOES NOT MEAN
THAT THE AUTHOR OR THE PUBLISHER ENDORSES THE INFORMATION THE ORGANIZATION
OR WEBSITE MAY PROVIDE OR RECOMMENDATIONS IT MAY MAKE. FURTHER, READERS
SHOULD BE AWARE THAT INTERNET WEBSITES LISTED IN THIS WORK MAY HAVE CHANGED OR
DISAPPEARED BETWEEN WHEN THIS WORK WAS WRITTEN AND WHEN IT IS READ.

For general information on our other products and services, or how to create a custom For
Dummies book for your business or organization, please contact our Business Development
Department in the U.S. at 877-409-4177, contact [email protected], or visit www.wiley.com/go/
custompub. For information about licensing the For Dummies brand for products or services,
contact BrandedRights&[email protected].
ISBN: 978-1-394-22104-2 (pbk); ISBN: 978-1-394-22105-9 (ebk). Some blank pages in the print
version may not be included in the ePDF version.

Publisher’s Acknowledgments

Some of the people who helped bring this book to market include the following:
Project Manager and Editor: Sales Manager: Molly Daugherty
Carrie Burchfield-Leighton
Sr. Managing Editor: Rev Mengle
Acquisitions Editor: Traci Martin

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Table of Contents
INTRODUCTION................................................................................................ 1
About This Book.................................................................................... 1
Foolish Assumptions............................................................................. 2
Icons Used in This Book........................................................................ 2
Beyond the Book................................................................................... 3

CHAPTER 1: Introducing Generative AI..................................................... 5

Solving Real Problems.......................................................................... 5
Seeing the Reality Amid the Hype....................................................... 7
Understanding LLMs............................................................................. 9
Employing Generative AI.................................................................... 11
Exploring the Use Cases..................................................................... 13
Creating........................................................................................... 13
Analyzing......................................................................................... 14
Governing....................................................................................... 14
Taking Care with Generative AI......................................................... 14

CHAPTER 2: Employing Generative AI in the Real World.......... 17

Turbocharging Marketing................................................................... 17
Supporting the Support Team........................................................... 19
Operationalizing Your Success.......................................................... 21
Powering Up Learning and Development........................................ 22
Enabling the Product Team................................................................ 24
Helping Human Resources................................................................. 25

CHAPTER 3: Getting Started with Generative AI............................. 27

Articulating Your Business Goals....................................................... 27
Mapping Your Use Cases.................................................................... 28
Knowing What You Need.................................................................... 30
Setting Expectations............................................................................ 31
Getting Quick Wins.............................................................................. 32
Choosing the Right Partner................................................................ 33

Table of Contents iii

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 CHAPTER 4: Ten Generative AI Success Factors............................... 35
Playing Key Roles in Success.............................................................. 35
Spreading Impactful Use Cases......................................................... 36
Putting People First in Change Management.................................. 37
Establishing Guardrails for Brand
Safety and Consistency....................................................................... 37
Setting Governing Principles.............................................................. 38
Becoming an AI-Ready Organization................................................ 38
Carefully Evaluating Vendors............................................................. 39
Building Ongoing Training.................................................................. 39
Shifting the Hiring Mindset................................................................ 40
Thinking Big.......................................................................................... 41

iv Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Introduction
M
any people are intrigued and surprised by how quickly
artificial intelligence (AI) has blown into the main-
stream. What seemed like a futuristic consideration is
suddenly something you fear you’re already behind on under-
standing and adopting. But as you race to keep up, you also may
fear the downsides some of those headlines have shared.

Take a breath and be ready for words of comfort. The media

attention has been sudden, but expertise in AI is already estab-
lished and ready for you to tap into. The crazy stories about risks
are real, but so are the AI safeguards available if you choose the
right expertise and technology.

The stories about the stunning power and remarkable usefulness

of AI are just as important, too. Your organization can benefit
from AI in astounding ways by implementing a full-stack gen-
erative AI platform. That’s how you uncover enterprise-wide use
cases for efficiently creating new content and transforming exist-
ing work, analyzing data to generate insights, and governing con-
tent to ensure compliance, accuracy, and brand consistency.

The best advice is to adopt technology that’s fully customizable,

generating content and answering questions about your organi-
zation’s data in ways that are totally aligned with your organi-
zation’s goals and values. Your generative AI technology must
integrate seamlessly into your workflows and existing tools, it
should function across the entire enterprise, and there should be
full transparency regarding how it thinks and operates.

About This Book

Enterprise Generative AI For Dummies, Writer Special Edition, is your
guide to quickly and safely moving into this new era. This book
helps you harness the power of generative AI and gain a competi-
tive advantage. Check these pages to learn about the state of gen-
erative AI and how enterprises benefit from its capabilities. You
take a deep dive into the large language models (LLMs) that make

Introduction 1

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 generative AI happen, the potential use cases, the challenges, and
the countermeasures. You get tips on how your organization can
articulate and prioritize its goals, how to get started, and how to
ensure success.

This book also covers Writer’s full-stack generative AI platform

that’s enterprise-grade, learns from your organization’s own
data, is fully customizable to fit into existing workflows, and is
compliant with Systems and Organization Controls (SOC) 2 Type
II, General Data Protection Regulation (GDPR), Health Insurance
Portability and Accountability Act of 1996 (HIPAA), and Payment
Card Industry Data Security Standards (PCI DSS).

Foolish Assumptions
When writing this book, I made some assumptions about you, the
reader:

»» You may be a marketing, sales, customer support, or

operations leader but you don’t have a lot of technical
understanding about AI.
»» You’re eager to learn more about the business value of
generative AI.
»» You’re looking for a primer into how your enterprise can
implement generative AI successfully, safely, and easily.

Icons Used in This Book

In the margins of this book, you notice some icons here and there.
They’re like a roadmap pointing out important spots:

Skip a bit of text if you must, but be sure not to miss the key
points shared next to this icon.

This book is intended to be full of actionable ideas, and this icon

points to them.

2 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 AI is powerful, but you need to watch for a few potential chal-
lenges, as noted next to this icon.

Beyond the Book

The 48 pages of this book offer a tantalizing overview into gen-
erative AI. You may come away with an appetite to learn more, so
visit writer.com for more info on generative AI, use cases, and
resources, including its full-stack enterprise-grade platform.

Introduction 3

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Using AI to solve problems

»» Comparing reality versus hype

»» Seeing how LLMs work

»» Investing in generative AI

»» Exploring use cases

»» Keeping ahead of any risks

Chapter 1
Introducing
Generative AI

A
rtificial intelligence (AI) has been the subject of fascinat-
ing science fiction for many years, but it has recently filled
the real-world headlines. Businesses of all types are dis-
covering just how many actual problems and challenges can be
solved through the use of AI. It’s not fiction, and it’s not hype; it’s
a real game-changing reality.

This chapter dives into that reality, explains how large language
models (LLMs) bring generative AI to life, explores the business
value and uses of generative AI, and ventures into the obligatory
cautionary tales. I share effective countermeasures for avoiding
trouble, too. One of the best bets is the opportunity to incorporate
an enterprise-grade, fully transparent platform with your orga-
nization’s business processes and existing IT architecture.

Solving Real Problems

Whatever industry you’re in, odds are you’ve been bombarded
with talk about AI. There’s no escaping it. It’s all over news arti-
cles, popping up in industry journals, and if you listen in on such

CHAPTER 1 Introducing Generative AI 5

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 things as the earnings calls of technology companies, the term
“AI” sounds like it’s on repeat. It’s as ubiquitous as pumpkin
spice flavors and scents in the autumn.

This book is packed with information about generative AI, but you
may be wondering whether it’s a good use of your time to find
out more. Amid all the hype, is it enough of a game-changer to be
worth your attention?

Or to put it another way, what can AI really do for your business?

What problems can it solve? Turns out that, unlike those pumpkin
spice flavors that are forgotten by New Year’s, AI really can make
a lasting mark on your business operations and growth. Here are
some of the headaches it can cure:

»» Production bottlenecks: I’m talking about processes that

are stuck and unable to keep up with the demands of
customers. Generative AI breaks through bottlenecks. It
automates processes, improves efficiency, and helps the
humans on your team make decisions that are not only faster
but also better. It can help you increase output, get the most
out of your resources, and speed up development cycles.
»» Tedious tasks: If your company is like most, you’ve been
having trouble finding and keeping good talent. People are
increasingly willing to hit the road if they’re buried in mundane,
repetitive work. Generative AI can uncomplainingly tackle
tedious tasks, freeing up your human brainpower for real
value-creating initiatives that your people will find fulfilling.
»» Inconsistencies and noncompliance: Maintaining content
that’s consistent across the organization and compliant
with internal and external standards is essential, and it’s a
total headache. As employees turn over and workloads
increase, it’s easy for discrepancies and errors to pop up
and multiply. Generative AI can identify these issues, offer
insights and recommendations, and even automatically fix
them. Small problems can be eliminated before they get
out of hand.
»» Training hurdles: You sure would like that new hire to get
up to speed quickly, but it takes time to gain the necessary
knowledge and skills. Generative AI can get new hires
moving more speedily by automatically generating content
such as training materials and job simulations. Personalized
instruction can fill knowledge gaps.

6 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Generative AI isn’t just a new flavor for your business operations.
It has the power to tackle real-world problems that you’re likely
facing. It’s worth reading on to learn more about how that can
happen.

Customers, too, can end up more satisfied when an organization

employs generative AI. When equipped with information-retrieval
solutions, the technology can answer questions quickly and accu-
rately and offer hyper-personalized recommendations. It can
handle some customer interactions entirely on its own and also
improve live human interactions by empowering agents, elevat-
ing service levels, and creating instant conversation summaries.

Over on the research and development side, generative AI can

help researchers explore potential new products and services. It
can analyze customer data and industry trends and suggest ideas
for innovation. Folks in marketing find generative AI useful for
creating product descriptions, emails, landing pages, FAQs, and
even ads. A well-curated LLM with appropriate AI guardrails can
allow all this content to be fully on-brand and totally accurate.

Seeing the Reality Amid the Hype

Seemingly endless possibilities exist for reducing or speeding up
work, tapping into powerful insights, and gaining competitive
advantages through the use of AI. It can sound like part hype and
part miracle, but you may be surprised to learn just how many
people and organizations are already sophisticated users of AI for
many different purposes.

In fact, most people’s daily lives are touched by the use of AI in a

lot of ways already, whether they know it or not. For example, if
you listen to Spotify or watch Netflix, you’re getting recommen-
dations and classifications that are informed by AI. Your Google
searches are, too. And the stuff that pops up on TikTok.

Indeed, you can use AI in far more ways than I could ever possibly
list, but the focus for this book is using generative AI. Generative
AI refers to not just making recommendations but actually creat-
ing new data or content or generating insights by using natural
language processing (NLP) and machine learning (ML).

That word generative means it’s not just recognizing or classify-

ing data but actually coming up with something entirely new by

CHAPTER 1 Introducing Generative AI 7

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 using AI — often something that sounds like it was created by a
human. As you’ve no doubt read in the headlines, the stuff created
by AI could be text, images, code, audio and music, or video and
animation. For the purposes of this book, though, the focus is on
text generation.

To be even more specific, this book focuses on using generative

AI for creating, analyzing, and governing text-based content in
business contexts. There’s a lot more detail on these use cases
later in this chapter, but it’s worth setting the stage here.

»» Creating is pretty much what it sounds like: using AI to come

up with something new. It also may mean editing or revising
something that has already been created, by a person or AI,
perhaps by turning it into a different format.
»» Analyzing means taking an in-depth look at content of some
kind and generating insights. That may mean spotting trends
or reaching conclusions of some sort, perhaps even analyz-
ing sentiment amid a batch of customer feedback.
»» Governing has to do with examining text with an eye for
whether it lives up to certain standards. They could be legal
or regulatory requirements, factual details related to
products, brand consistency, or inclusivity.

The kind of work performed by technology can have tremen-

dous implications on your business operations, including human
resources. For example, if you’re losing employees to burnout or
overwork, you know that turnover comes at a very high cost, both
in terms of recruiting and replacing workers, and moving forward
without their institutional knowledge.

What’s more, labor shortages have made it hard to fill every job,
and AI can be a handy force multiplier. But that brings up a super-
important point. A lot of the hype you may have read includes
worries about AI replacing people. That sounds threatening, but
it isn’t really an accurate description of how AI may impact the
humans on your payroll.

The reality is tapping into the assistance of AI can improve the

employee experience and take tedious tasks off workers’ plates.
That makes room for upskilling and reskilling and engaging
employees in satisfying value-adding work. Rather than threat-
ening jobs, AI holds the potential to improve them.

8 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Understanding LLMs
AI relies on building large language models with the help of
machine learning. But what does that actually mean and how does
it work?

ML is in a lot of ways similar to human learning, and an LLM is

learning language kind of like a human does. An LLM in train-
ing is exploring the language to find connections and statistical
patterns in the ways that words are used. It’s immersing itself in
the language, and as it recognizes patterns, it figures out how to
predict them.

Ultimately, the LLM is learning how to communicate and under-

stand language, so it can respond to a question or request by
generating an answer formed in natural language. But it’s worth
digging more deeply into those three letters, L, L, and M.

»» The first L, for large, is an important point. It takes vast

amounts of data to train an LLM. Immersion implies a lot of
data. That data may include vast volumes of information
from the Internet or a more specific, tailored set of data.
»» The second L, for language, is what it’s being immersed in, of
course. It’s learning pattern recognition, figuring out
communication in the same way humans do.
»» The M is for model. A model is a type of design, and when it
comes to LLMs, there are different kinds of models that are
created for different purposes.

Most of the headlines people have seen in recent years have

been focused on general, consumer-use models such as GPT and
Google Bard. But there are other models that are designed spe-
cifically for business use and for specific industries. The Palmyra
family of LLMs is an example.

Learning a language is only part of what humans do as they listen

to their parents and friends, pay attention in school, and watch
TV. They also learn substance, because what good is language if
you don’t have something to say?

The same is true about LLMs. They master the art of using lan-
guage from the data that trains them, but they also learn the
information contained in those data sets. To a user, an LLM can
seem like a genius, but the fact is, it doesn’t know everything.

CHAPTER 1 Introducing Generative AI 9

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Indeed, just like a human, an LLM is only as smart as the data it’s
trained on. It’s important for data to be as current as possible, and
critical that it be accurate.

It’s not hard to understand why those two factors are so impor-
tant. If the training data isn’t current, there may be new knowl-
edge the LLM doesn’t have and therefore can’t include in the
content it generates. And any inaccuracies included in training
data could end up in the content that the LLM generates.

But even that’s only part of the problem. AI has the potential to
perpetuate and amplify any biases and stereotypes that may show
up in training data. In doing so, it could stir up ethical issues,
damage reputations, even pose potential legal risks.

AI bias happens when algorithms, models, and datasets have

assumptions built-in that might lead to inaccurate or unfair con-
clusions. There are different kinds of bias that might pose trou-
ble, including historical, sampling, labeling, and confirmation
bias. Here are some real-life examples (I’ve left out the names
of the companies involved, but these are actual things that have
happened).

A decision-making algorithm designed for law enforcement was

found to falsely flag Black defendants as potential future crimi-
nals twice as often as White defendants. The same tool was also
more likely to inaccurately portray White defendants as low-risk.
This is known as historical bias.

A real-world example of a sampling bias was a hiring tool that was

trained on years of résumés from past candidates. Unfortunately,
most of the candidates were men, and the algorithm ended up
favoring male candidates.

What about labeling bias? Here’s a horrifying real-world example:

An early photo-labeling algorithm was found to tag Black people
as gorillas.

And confirmation bias happens when data is mistakenly inter-

preted to confirm an existing belief. This can happen in the world
of social media, where algorithms are more likely to feed you facts
that match and amplify your worldview — even if your worldview
includes inaccurate conspiracy theories.

10 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 If all this makes you nervous, that’s fine, but please don’t close
this book and think that AI is too risky. These kinds of nightmare
scenarios don’t have to come true. LLMs don’t have to behave like
a Wild West of language and knowledge, and smart enterprises
know how to make technology work safely and intelligently.

One of the biggest keys is transparency. Some of the popular con-

sumer AI tools that are all the rage are like a black box into which
massive amounts of data have been poured and stirred up. You
don’t know what you’re going to get out of them because you
don’t know what was put into them. It’s like going to the grocery
and buying some super-processed food that has no label showing
ingredients.

With AI, the solution is to partner with LLM and generative AI

vendors who are completely transparent about training data
sources and the various cutting-edge algorithms and technolo-
gies employed. There are many approaches for increasing trans-
parency, accountability, and explainability.

Your enterprise needs a partner whose sources and models can be

audited for toxicity and other factors that can lead to poor out-
puts. There should be privacy mechanisms to ensure that insights
can be obtained without exposing individual data points. You’ll
want to be certain there are techniques that monitor for regula-
tory compliance.

In short, you can feel much more confident in your AI solution

if there is no black box, but rather an open window that lets you
know what’s happening inside. It can certainly seem like magic,
but in this case, you need a magician committed to sharing the
secrets of how the magic works.

Employing Generative AI
Generative AI may have hit the mainstream headlines with
the sudden consumer popularity of ChatGPT, Google Bard, and
CoPilot, but smart enterprises have already been busily employ-
ing the technology big-time. Practically every industry — from
healthcare to financial services to retail and a whole lot more —
has companies investing in generative AI tools.

CHAPTER 1 Introducing Generative AI 11

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 These AI pioneers are experimenting with technology and rolling
it out across IT, marketing, sales, operations, and support teams.
The human resources team is a growing user, as are learning and
development, brand management, research and development,
and even legal teams.

This means a lot of professionals are already clued in to the sig-

nificant benefits. Polling suggests that they also have at least
some insights into the risks, such as data privacy concerns and
the potential for inaccuracies.

Writer conducted a 2023 survey on the topic of generative AI in

the enterprise. The survey gathered input from 466 leaders at
the director level or above at companies with at least a thousand
employees. Here are some of the findings:

»» Fifty-nine percent of companies have purchased least one

generative AI tool or have plans to do so soon. At some of
the companies not into generative AI yet, individual employ-
ees are finding ways to use it.
»» Only 7 percent of respondents said it’ll never happen at their
company, and as far as they know, no one is using genera-
tive AI on their own.
»» Nineteen percent are already using five or more generative
AI tools.
»» Respondents are using generative AI in pretty much every
function across the enterprise.
»» Fifty-six percent of respondents believe generative AI is
boosting productivity by 50 percent or more.

Figure 1-1 also gives you some details on how companies are using
generative AI, according to the Writer survey. When asked what
the top three generative AI use cases were in their companies,
Figure 1-1 shows those results.

The survey also turned up one finding that’s a bit concerning,

though not entirely surprising given the generative AI hype in the
media. The most often used tool is ChatGPT — some 47 percent
of respondents said their companies use it, and of regulated com-
panies, the share is 52 percent. Ironically, ChatGPT is also the tool
most frequently banned from use. An issue with using ChatGPT
in the enterprise setting is that organizations are opening them-
selves to security, privacy, brand reputation, and governance risks.

12 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 FIGURE 1-1: How companies are using generative AI.

Exploring the Use Cases

Many different use cases exist for generative AI. It’s possible
to boil down the most popular uses into a few buckets. As you
continue through the pages of this book, you see repeated refer-
ences to how AI delivers on use cases that fit into three general
categories.

Creating
Generative AI is about creating something new — generating
high-quality, accurate output. Text content may be the most com-
mon use, and that includes articles, reports, emails, social media
posts, ads, and so on. AI may take recordings and turn them into
derivative content in numerous formats, ensure web content is
SEO-ready, or adapt press releases to multiple audiences.

But the possibilities don’t stop with text. AI can generate com-
puter code, for example. It may autocomplete code that a developer
starts or come up with code based on a request from a developer.
And AI tools create images, too, either by adapting and altering
existing images or creating new ones, based on a textual request.
These tools can handle graphic design requests and design logos.
AI can paint paintings, design sculptures, create gaming charac-
ters, and even write music.

CHAPTER 1 Introducing Generative AI 13

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 After something is created, there’s always room for improvement.
AI can lend a hand with that, too, editing to refine or improving
existing work. Or it could transform an existing work, such as
creating new versions to adjust for industry, function, persona,
or business context. The create use case can also take the form of
repurposing existing works into different formats, such as adapt-
ing a video into a blog post or a white paper into an email.

Analyzing
AI tools can analyze the key themes, mood, or sentiment that
comes through in content and get answers in seconds. The tech-
nology can spot patterns in large sets of generated data and review
content with an eye toward quality analysis. It can analyze and
synthesize data to surface insights and spot trends and answer
questions, including queries that tap into your own company’s
knowledge and data.

As part of the analyze use case, AI can compile and summarize

reams of information, proprietary to your organization, into
research briefs. It can create summaries of documents, webpages,
and recordings, too.

Governing
The govern use case includes a focus on compliance, looking for
language that runs afoul of legal and regulatory rules. It finds
incorrect terminology and statements and works to prevent data
loss and global compliance problems. This type of AI work also
means checking for factual accuracy, detecting claims that are
wrong and suggesting replacement wording.

The AI can also police for brand consistency, making sure that all
work reflects the corporate brand, messaging, and style guide-
lines. And it can ensure that language used is inclusive and
unbiased.

Taking Care with Generative AI

Despite all the amazing things you can accomplish with genera-
tive AI, there’s still a catch. Yes, you need to be mindful about
some concerns, constraints, and potential risks. The important
thing is to be aware of them and understand the countermeasures.

14 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 After all, you face risks every time you climb into a car, too, but
you employ guardrails such as fastening your seatbelt. Just look at
the AI technology in your car — it augments and makes the expe-
rience safer, but you’re still in control. You drive on, mitigating
the risks and moving forward. That’s what you do with generative
AI, too.

As AI becomes more powerful, complex, autonomous, and

ubiquitous, guardrails become all the more important. Guardrails
help ensure AI is being used responsibly and ethically, and not
misused for nefarious purposes. Guardrails protect against bias
and work for fairness. They help ensure compliance with legal
and regulatory requirements.

Guardrails also maintain a connection with humanity. They

facilitate human oversight and help reinforce the idea that AI is
a tool rather than a replacement for humans and their ability to
make decisions. And knowing that guardrails are there helps reg-
ular humans build confidence in AI.

Complementing the assurances of guardrails are the counter­

measures you should employ to fight back against various specific
concerns. Take generative AI hallucinations (sometimes known as
plausible BS) as an example. On occasion, a generative AI tool cre-
ates content that sounds credible and plausible, but it’s inaccu-
rate. Hallucinations are why some companies have banned the use
of generative AI technologies targeted toward individual users,
such as ChatGPT.

So, you take countermeasures. Fact-checking is an obvious one

for your content regardless of whether it’s been written by a
human or an LLM. You may not be surprised to learn that AI-
powered fact-checking tools already exist. You can also employ a
claim-detection tool that automatically flags statistics, facts, and
quotes that people want to verify.

One of the best countermeasures is ensuring that your LLM is

trained on clean, accurate, curated data, with the help of a vendor
committed to full transparency. For example, if your generative
AI is basing its outputs on your own company materials and other
data that you know is accurate, fair, robust, and reliable — rather
than a broad crawl of the web — you tremendously reduce your
risk of hallucinations.

CHAPTER 1 Introducing Generative AI 15

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Other concerns related to generative AI include the following:

»» Data security and privacy: Consider the risks of using tools

built on LLMs designed for consumer use, which reserve the
right to keep, access, and use your data. If someone uses
confidential or customer data in a prompt, it becomes part
of the LLM database, which can put intellectual property and
other sensitive data at risk.
»» Copyright: Much ongoing debate exists about the use of
copyrighted materials in LLM training. Creators whose work
ends up teaching an LLM have been asking for more
protections, so this is an area of active litigation. Folks may
also debate whether AI-generated content is protected
under copyright law. According to the United States
Copyright Office, that depends on the level of human
creativity, but unedited content created by the most popular
AI systems probably doesn’t qualify for protection — and
that means if your organization wants to sell AI-generated
material, that could be an issue.
»» Compliance: You expect your human content creators to
comply with corporate guidelines, branding norms, styles,
terminology guidelines, and regulatory requirements. A
general-purpose tool won’t know all the rules, so full
compliance isn’t a given.

The countermeasure to these risks is an enterprise-grade, custom-

izable generative AI platform, complete transparency, thoughtful
guardrails, and when applicable, the addition of human creativity.
You want a platform that’s tailored to your needs, that’s certified
under privacy and security standards that govern your organiza-
tion, and that won’t store data or content any longer than it’s
needed.

16 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Giving marketing a hand

»» Offering support to the support team

»» Upping the game for operational leaders

»» Teaching new tricks to learning and

development

»» Powering up product development

»» Helping out human resources

Chapter 2
Employing Generative
AI in the Real World

I
f you’ve decided to up your organization’s generative artificial
intelligence (AI) game, before you go shopping for a solution,
you should gain a good understanding of your needs and how
generative AI works in real-world settings. This chapter spells out
how AI can fit into the workflow of a half-dozen different teams
within your organization, what challenges it can solve, and what
tasks it can tackle. It offers real-world examples of how genera-
tive AI helps to create, analyze, and govern your content.

Turbocharging Marketing
Your enterprise may have the most mind-blowing products or
incredibly essential services out there, but without the marketing
team, you’ll spin your wheels in frustration. You can’t succeed
without connecting with customers to understand their needs and
let them know how you’re going to solve them.

CHAPTER 2 Employing Generative AI in the Real World 17

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 At the outset, these players conduct market research to get a
sense for consumer needs and preferences. They craft compel-
ling advertising campaigns. They manage the brand’s image and
reputation. They do their work in platforms ranging from digital
media to traditional advertising, from webpages and catalogs to
blog posts and countless other approaches. They make sure your
company’s voice is heard in a competitive landscape, and if all
goes well, they drive sales growth.

To do this well, it requires a lot of different kinds of effort, per-

haps more than seems humanly possible. To begin with, there’s a
whole lot of content to create, in multiple different formats. The
more people involved in that, the greater your chances of straying
off-brand. It’s even more of a potential issue if you’re involving
partners down the sales chain who aren’t as directly connected to
the organization.

An enterprise, generative AI platform can give your market-

ers a huge advantage. With AI, they can create a whole lot more
content. They can bring products and services to market faster,
employing distribution shortcuts. And they can be sure the whole
enterprise stays on-brand. Here are some examples from each of
the primary use cases of generative AI:

»» Create: Your AI tool can write the first draft of an ebook

about the product. Think of how much of your marketers’
time that would free up.
»» Analyze: Ask the AI platform to process a webinar recording
and summarize the key takeaways.
»» Govern: Run some advertising copy through the platform
and have it flag spots where it gets terminology wrong,
makes incorrect statements, or violates legal or regulatory
standards.

From the perspective of the marketing team, the functional

requirements of generative AI include knowing your products,
speaking in your voice, writing in your organization’s style, inte-
grating with all necessary sources of data, and being able to detect
claims and check the facts.

18 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 CASE STUDY: MARKETING
THE BRAND
A technology company was growing at a healthy pace but finding the
competitive environment to be increasingly challenging. It needed to
get far more efficient and effective in its marketing approaches, scal-
ing its content output without adding any more human resources.

Generative AI helped that marketing team boost its writing output by

50 percent, with greater brand consistency and more personalized
content. It took a hundred blog posts and rebranded them to better
align them with the brand style. It made its product documentation
more conversational, which again was better aligned with the compa-
ny’s style and voice. And it was able to verticalize content so that it
would cater to specific industries and personas.

Supporting the Support Team

Your organization won’t last without customer loyalty, and
customer support team members are essential in maintaining
satisfaction and building loyalty. They’re often the first point of
contact for customers needing help, they address questions, they
resolve issues.

Your support team members gather feedback, which is a goldmine

for improving products and services. And they help customers get
up and running with a product, talk up product updates, handle
troubleshooting. They’re the key to smiling customers.

This is a crucial job, for sure. And not an easy one. They need to
have a ton of knowledge or have a fast-and-easy way to find that
knowledge. AI can build the knowledge base that support team
members tap into. Your support employees also need to be able to
make sense of their customer interactions, or else you’ll never be
able to spot opportunities for improvement. AI can help with the
analysis of interactions. And like your marketers, your support
team members need to reflect your brand’s personality and tone,
and AI helps flag communications that stray.

CHAPTER 2 Employing Generative AI in the Real World 19

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 What are other specific things these teams might ask of an enter-
prise AI tool?

»» Create: The support team can have generative AI write

a knowledge base article on the latest product feature.
»» Analyze: A generative AI platform can scour customer
support survey responses to come up with insights on
areas of improvement to consider.
»» Govern: The system can suggest support responses to
customer inquiries that are fully compliant with brand
standards.

To sum up the functional requirements your generative AI tool

must have from the point of view of the support team, it has
to know your products inside and out. It must be able to offer
answers and do so in the voice of your brand. It has to recognize
claims and verify the facts in those claims. It must be able to add
to the knowledge base and create FAQs. And it has to ingest any
kind of data format to gather customer insights.

CASE STUDY: FILLING THE

KNOWLEDGE BASE
Accurate and efficient customer support requires a knowledgeable
team, but nobody can know everything. That’s why the knowledge
base is so essential, making support team members instant experts in
whatever the topic at hand might be. But who keeps the knowledge
base filled with knowledge?

One tech company faced a daunting task of scaling up its support

knowledge management. The team of writers creating knowledge
base articles needed to grow, but the number of articles had to grow
even faster. A generative AI platform helped in multiple ways.

Trained on style guides and content guidelines as well as organiza-

tional and product info, the AI system helped ramp up new writers
quickly and allowed all writers to work far more efficiently. It also
enabled a team-publishing model that let even non-writers produce
great content. The result: The support knowledge management team
quickly went from publishing 200 articles a year to creating more than
a thousand.

20 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Operationalizing Your Success
At the heart of day-to-day activities is the operations team
with a broad range of functions on its plate. All processes from
design to production to delivery are in its sights, along with
operational planning and forecasting, finance and strategy,
quality control, and the supply chain management that ensures
smooth sailing.

Operational leaders will want in on the productivity gains prom-

ised by AI, as they work to ensure operations are well-coordinated
and programmatic. They’ll find generative AI useful in overcom-
ing hurdles in multiple areas they control.

For example, sluggish development time can put a crimp in com-

petitiveness, but generative AI can save time on such elements as
product description creation. A robust program can help strate-
gic planners establish more cogent objectives. It can help investor
relations wade through stacks of press releases to prepare meet-
ing notes for earnings calls. A few more examples:

»» Create: Just watch as generative AI automatically creates

product descriptions as details hit the product database.
»» Analyze: A generative AI platform can evaluate quarterly
reports to gain insights for developing objectives and key
results.
»» Govern: Investor updates have notoriously stringent
regulatory requirements, and accuracy is especially vital
here. Generative AI keeps them in line.

From the perspective of operations, the functional requirements

of AI include knowing products inside and out, fully integrating
with the tech stack, ensuring writing is based on business con-
text, checking facts for all claims, and ingesting data from all for-
mats for understanding and analysis.

CHAPTER 2 Employing Generative AI in the Real World 21

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 CASE STUDY: IMPROVING
OPERATIONS
A retailer in a competitive fashion sector was seeking ways generative
AI could boost efficiency and improve operations. It found answers in
many places.

For example, a product can’t hit the online market until product
descriptions are ready, but those take time. The retailer cautiously
tested to ensure AI-produced descriptions would achieve conversion
rates that match human-produced descriptions. The answer was yes,
and AI got products posted and selling more quickly.

Just as important, sending that work to AI freed up valuable

people resources for areas with higher impact. Letting AI handle
the mundane meant creative resources could focus on building
new competitive advantages.

Powering Up Learning and Development

Organizations rely on their learning and development (L&D) team
for fostering ongoing employee growth and corporate success.
The team works day in and out identifying skills gaps and creat-
ing training programs to fill the gaps. That may be in-person or
online classes, workshops, or mentoring, but in all cases, meas-
urable growth in performance and productivity is a must.

Ultimately, that requires a whole lot of content creation, in a wide

range of formats. It can be overwhelming for the team to keep up
with that need, especially in a growing organization where L&D
staffing levels aren’t keeping pace, a common dilemma. Some
of the required work involves brand-new creative efforts, while
much is mundane repurposing of existing materials into different
formats.

Generative AI can fill in those gaps and meet the challenges. It

can customize existing training materials by team or department
function. It can write a video script based on a training manual,

22 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 or vice versa. It can analyze a skills gap and determine ideal
resources for addressing it. Here are some other thoughts that fall
into our general buckets of AI uses:

»» Create: A generative AI platform can come up with a

best-practices article targeted at newly promoted managers.
»» Analyze: The platform can come up with some FAQs by
analyzing and categorizing what’s in an internal wiki.
»» Govern: In training materials it creates, as well as existing
materials, a generative AI platform can ensure compliance
with industry certification requirements and other vital
standards.

The functional requirements that your L&D team will demand

from generative AI include a strong grasp of the company’s cul-
ture and business context, and the ability to write in the com-
pany’s tone for whatever audience it’s targeting. It needs to be
able to ingest any data format and understand what it’s ingesting,
and it must be able to detect claims in order to check the facts in
those claims.

CASE STUDY: STREAMLINING

TRAINING
A growing IT security company was having trouble keeping up with
the expanding demands on its L&D team. It wasn’t just a matter of
cranking out a high volume of content. There was a need to quickly
bring new writers up to speed on style guidelines and enforce consis-
tency, and ensure all materials were meeting compliance
requirements.

The company turned to generative AI to create written recaps of train-

ing videos. Its platform scoured existing training materials with a keen
eye toward brand, proper terminology, and compliance require-
ments. Its focus on consistency helped with a globalization push, in
part by better facilitating translation. And AI was easy to adopt, with
the platform allowing revisions right in the existing content manage-
ment system.

CHAPTER 2 Employing Generative AI in the Real World 23

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Enabling the Product Team
Product leaders are developing products and services, but in a
broader sense, they’re creating hopefully delightful user experi-
ences. The product team requires not only technological skills for
bringing concepts to life but also a keen ability to process and
analyze market research and customer feedback, create commu-
nications to facilitate successful adopting, and efficiently docu-
ment product updates and fixes.

That’s a whole range of talents that more than likely doesn’t all
exist in any individual person. Keeping up with customer needs in
a competitive environment requires constant vigilance, the ability
to spot trends, and a continual eye on potential feature improve-
ments. Generative AI can help pull together disparate sources of
input for those purposes. And any logjam in the development of
supporting content, such as release materials and FAQs and prod-
uct error messages, can delay product introduction. An AI plat-
form can ease those logjams, too, through content generation.

Here are some things product developers may ask of AI:

»» Create: A generative AI platform can ingest and crunch a list

of features and bug tickets to come up with release notes.
»» Analyze: The platform can study customer feedback to find
insights for what new features to prioritize.
»» Govern: The right platform can work right inside product
design tools to police how the brand is represented, monitor
appropriate product terminology, and suggest inclusive
language.

The functional requirements that the product design team will

demand from generative AI include a full knowledge of the com-
pany’s products and terminology, the ability to write in a tone
that matches the brand, a watchful eye for accessible and inclu-
sive language, and the ability to ingest and understand all data
formats while also working inside existing tools.

24 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 CASE STUDY: AI IN THE
DESIGN PROCESS
A well-known financial technology company is constantly developing
new and improved products. These products need to meet customer
needs and wishes, of course, but the products and any supporting
materials also must adhere to brand and style guidelines, and inclu-
sive language.

This company used generative AI to build these important sensibilities

right into its design systems. Its platform brought the style guide right
to the place where designers do their work, ensured proper terminol-
ogy, and privately flagged non-inclusive language to promote healthy
but discreet learning opportunities.

Helping Human Resources

Ensuring a positive and effective work environment is the ulti-
mate aim of the human resources (HR) team. Its professionals
recruit and onboard new employees and work to retain the top
talent. HR creates and oversees policies of all kinds, covering
employee relations, benefits, compensation, performance man-
agement, and other areas.

HR work comes with its share of challenges. Keeping job descrip-

tions up-to-date, for example. Compliance is a big one, too, with
labor-related laws and regulations that vary from place to place
and change often. Maintaining a diverse and inclusive environ-
ment is a vital challenge, as well. Generative AI will help write and
maintain HR-related content, ensuring it’s aligned with company
policies, inclusivity goals, and regulatory requirements. Here are
a few general examples:

»» Create: Generative AI can write up a job description and

make sure it’s doing so in inclusive language.
»» Analyze: AI can listen to a recording of a job interview and
create a summary.
»» Govern: AI can flag non-inclusive language in employee
communications, then make suggested revisions.

CHAPTER 2 Employing Generative AI in the Real World 25

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 The functional requirements that the HR team will have for an AI
platform include full understanding of the company, its termi-
nology, and the kinds of job roles on the payroll. It must always
use accessible and inclusive language that reflects the voice of the
company. And it must be able to tap into and analyze all different
forms of data.

26 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Establishing your goals

»» Mapping your use cases

»» Discerning your needs and expectations

»» Succeeding quickly

»» Choosing a partner

Chapter 3
Getting Started with
Generative AI

T
his chapter explores how you can expect a generative artifi-
cial intelligence (AI) program to tackle your organization’s
needs, what kinds of goals to set, how to decide what use
cases to implement, what kinds of results you may expect, how to
move your organization in this direction, and ultimately, how to
make the choice of a partner.

Articulating Your Business Goals

With all you read about AI these days, it’s easy to get the feeling
that your organization may already be behind. Given how quickly
generative AI went from being a curiosity for early adopters to a
useful technology that truly anyone can access, you can be for-
given for having an itchy trigger finger.

That said, you shouldn’t plunge into AI just for the sake of doing
so. You may gain a cool marketing hook, but what else will it do
for your enterprise? As with any other investment or directional
shift, take a breath and a step back to really articulate what you
can achieve in this space.

CHAPTER 3 Getting Started with Generative AI 27

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Chapters 1 and 2 detail the many useful things AI can bring to your
organization — now’s the time to determine where it will fit in
for you. What exactly are your enterprise’s goals that’ll be aided
by a generative AI program? What are the challenges or problems
you’re trying to address? Where are you feeling the need to get
ahead?

This is where time-honored concepts of process improvement

and innovation are still very much applicable. Some key steps
should be followed:

1. Define your purpose.

Do you hope to improve productivity? Offer new and innova-
tive services? Facilitate operations? Create new products? Solve
complex problems? Enhance customer satisfaction? Raise the
bar on employee engagement and retention?
2. Write down your objectives.
Take the answers to the questions in Step 1 and translate
them into detailed use cases and business value. These may
describe how generative AI automates repetitive processes,
cuts operational costs, and reduces human error. Your
objectives may detail how you improve customer service or
the employee experience, how you gain analytical insights
into customers and operations, how you improve compliance
and increase inclusivity. There are plenty of challenges that
may get a boost from generative AI.
3. Establish measurable goals.
Set what are known as SMART goals: specific, measurable,
achievable, relevant, and time-bound. How can you know if
you’re making an impact if you have a fuzzy, nonspecific aim
like “generate more content and do so a whole lot faster”?
You’re better to create a quantifiable goal to reach by a
certain time — something aspirational but also attainable.

Mapping Your Use Cases

The process known as use case mapping is a great way to put some
meat on the bones of your objectives and SMART goals. AI use
case mapping examines specific use cases within your organiza-
tion and figures out the potential benefits of introducing AI, as
well as how difficult doing so will be.

28 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 AI use case mapping has two primary dimensions:

»» Value: This is the specific outcome your organization will

achieve. Here’s where you calculate a return on investment
(ROI) that considers such things as how many people will be
positively impacted (the more the merrier, of course). Your
ROI may also calculate time to be saved because if you can
cut out the time your teams spend on low-value tasks, you’re
freeing them up for higher-value work.
»» Complexity: How easy or hard will it be to implement AI for
this particular use case? Maybe you’ve already got products
in place with AI capabilities you simply have not yet switched
on. Or maybe you need to add something new to the tech
stack. As you gauge complexity, you’ll also need to consider
integrations and how third-party products fit in.

Remember that this is called use case mapping. That implies com-
ing up with something visual or map-like. Check out Figure 3-1
for an example of how your organization may map out potential
use cases and compare which ones offer the most value or are the
least complex to implement. Generally speaking, the strongest
initial use cases are the ones that offer high value but are lower
in complexity.

FIGURE 3-1: AI use case mapping.

CHAPTER 3 Getting Started with Generative AI 29

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 The map offers direction, but the final decisions must involve peo-
ple. It’s vital to consult with business leaders in the areas where
implementation is being considered. They have key insights into
how ready their people are for adopting AI-powered solutions,
they know how it might fit into workflows, and they will have
insights into what the end result must look like.

Knowing What You Need

Being able to fully customize a large language model (LLM) is
an absolute must-have for an enterprise AI solution. Customiza-
tion is essential for harvesting all the benefits and mitigating the
risks — such as hallucinations and inaccurate content as well as
data privacy issues — that come from tapping into an LLM full of
who-knows-what information.

With the right solution, your organization can have its own LLM,
trained and fine-tuned with your data and content, along with
any other content that’s trustworthy and aligns with your needs.
It can connect to your existing data sources, including knowledge
bases, wikis, cloud storage, and chat channels. And transparency
is essential, so you know how the AI model is trained and how it
functions.

You may prefer to have this proprietary LLM hosted by your AI

solution provider. That’s a smooth solution. But depending on
your needs and your industry, you may need to own and host your
customized LLM. A self-hosting option is a must-have for some
organizations operating in highly regulated industries such as
healthcare and financial services.

Integration is another must-have. Integrating into your existing

data sources ensures the most effective customization. Beyond
that, your AI tool needs to integrate with your business work-
flows, existing tools, and processes. Expect your platform to be
accessible with in-line commands wherever people work.

One more must-have is enterprise readiness. Generative AI offers

benefits across the enterprise, but you’re only going to be able
to get the most out of them with a full-stack enterprise solution.
Enterprise-readiness maximizes customization and integration
and provides the privacy, security, regulatory compliance, and
administrative controls your enterprise needs.

30 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Look for such enterprise-friendly features as single-sign-on
access control, role-based permissions, multi-factor authentica-
tion, domain discoverability, activity audits and reports, and that
kind of thing. Expect enterprise-grade reliability, with real-time
and historical platform status, and a solid uptime commitment.

Every organization must be extremely careful with data, whether

it’s customer information or intellectual property or just about
anything else that’s been collected. But some verticals or geo-
graphic locations have very specific compliance requirements.

Healthcare organizations, for example, need to be ever watch-

ful when it comes to patient information, lest they run afoul of
the Health Insurance Portability and Accountability Act (HIPAA).
Service organizations often have an interest in System and Orga-
nization Controls (SOC 2). Those dealing with payment cards pay
close attention to payment card industry (PCI) security standards.

Therefore, expect compliance with whatever standards that mat-

ter to your organization. And when it comes to data privacy, an
enterprise will want an AI tool that specifically does not retain
and use your data or claim ownership of it.

Setting Expectations
There are, no doubt, specific key performance indicators (KPIs)
you want to impact. That’s the way to gauge what kind of return
you’re getting on your generative AI investment. That said, the
KPIs you’re seeking to move are specific to your own enterprise.
A common one is an increase in writing output for the marketing
department. What’s a reasonable expectation? Some users have
gotten pretty phenomenal results, such as a 50 percent increase
in output.

Perhaps you measure your productivity not in output but in hours

spent on tasks such as writing or percent of coverage of areas
where you may not have had the bandwidth to cover before. You
could seek a certain percentage reduction in writing time or cer-
tain number of fewer hours spent per month on a specific task
such as podcast content creation. What’s achievable? One com-
pany cut its content creation time by two-thirds.

CHAPTER 3 Getting Started with Generative AI 31

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Maybe generative AI is your answer for upping your customer-
support game. In that case, you may aim to increase your output
of knowledge management articles. One enterprise AI user saw its
team ramp up article production from under 200 articles a year to
more than a thousand.

Getting Quick Wins

Content creation isn’t just about implementing a powerful solu-
tion. Your team must know how to get your AI platform to do
what you need it to do. The amazing thing is that you don’t need
to learn new programming language or anything techie like that.
You should be able to communicate with your generative AI tool
in natural language, asking it questions or providing instructions.
These are known as prompts, and the key is including the right
info in your prompt.

Your generative AI platform delivers best when your prompt

answers important questions, such as who communicates to
whom, what is the format and length, where will it be communi-
cated, why does the audience need the content, how should it be
structured, and from what source material is it created? Here’s an
example of a specific prompt:

Write a 300-word LinkedIn post about the benefits of genetic

testing, based on this blog post from the ALS (Lou Gehrig’s
Disease) Association (here’s where you insert a hyperlink to
the source material). Convince people under 50 with a familial
history of ALS to get tested in the next six months. Use an
urgent but comforting tone.

You can start winning quickly with generative AI. Installing

a plugin brings AI-powered assistance into Microsoft Word,
­web-based applications, Outlook, Figma, and elsewhere. ­Writing
suggestions start showing up right out of the box, and they can
then be customized with your preferences, from punctuation
to number formatting to inclusive language, and more. H ­ aving
a consistent plug-in is important for maintaining corporate
standards — instead of having each application use its own gen-
erator and set of AI guardrails.

32 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 You can then incorporate company terminology and word lists.
You can even add frequently used content snippets to save time.
For example, if you have a company or product description that
you repeatedly use, add it as a reusable snippet. This makes it
available to your entire organization. And you can publish your
style guide to a password-protected location for creators to access.

Choosing the Right Partner

Generative AI holds great potential for your organization, and
the technology is evolving and innovating quickly. How will you
choose a partner for this journey? What questions should you ask
to settle on the right full-stack generative AI platform for the
needs of your enterprise?

As you research vendors, you should cover a wide range of topics.

Here are some thoughts:

»» Technical architecture and deployment: What’s the

vendor’s foundational LLM technology — open-source,
wrapper, or proprietary? What are the infrastructure needs?
How are data separation and secure processing managed?
Can the product integrate with multiple LLMs, or those
provided by clients?
»» Data life cycle management: Where does the data used to
train the foundational model come from? How does the
vendor prevent customer data from informing the broader
model? How is client data used for customization and
fine-tuning, and is it shared? How is personal data protected,
and sensitive data redacted?
»» Customization and integration: Can the LLM be fine-
tuned with private datasets? Does the product seamlessly
integrate with third-party services or apps that enterprises
commonly use?
»» Enterprise security: Does the product comply with stan-
dards such as SOC2, HIPAA, or GDPR? What authentication
methods are supported, and is single-sign-on supported?
Who can access foundational models? How are malicious
actors prevented from injecting harmful prompts?

CHAPTER 3 Getting Started with Generative AI 33

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 »» Scalability and performance: How does the solution deal
with large datasets and high-demand scenarios? How does it
handle scalability? And is there human oversight? How does
the solution address hallucinations?
»» LLM output compliance: How does the product mitigate
bias and inappropriate content? What about toxicity
detection — any industry standards or benchmarks for that?
What are the sources of bias data? How does the solution
ensure diversity of LLM-generated output?
»» Legal and regulatory compliance: How does the product
ensure that outputs don’t violate any intellectual property
rights? Who owns the data inputs and generated outputs?
Have the models been independently reviewed? Are there
compliance standards for protecting personal information?
»» Monitoring and reporting: How transparent is the product —
what tools and features offer insights into the model’s decision-
making? What visibility is there into telemetry and security
events? Are there reports regarding the accuracy of generated
outputs, or the effectiveness of controls?
»» Financial and operational considerations: Does fine-
tuning or customization cost extra? What kinds of training
and support are available? Can certain generative AI features
be disabled at an enterprise level?

34 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 IN THIS CHAPTER
»» Building your teams, your guidelines, and
your guardrails

»» Becoming AI-ready and spreading the

success

»» Adapting your hiring and training

mindset

»» Empowering your people

Chapter 4
Ten Generative AI
Success Factors

Y
es, this is a book about technology. But amid those bits and
bytes of information, it’s really a story about people. It
takes the right people to achieve generative AI success — it
requires choosing them, helping them adapt to change, ensuring
they know what they’re doing and why, collaborating more, and
finding the right people to help with your journey. This chapter
shares ten success factors for implementing a successful program
that works for your organization, as well as your people.

Playing Key Roles in Success

For any project to be successful, you have to make sure that you
have the right resources. The success of the initiative is driven by
a number of critical roles:

»» Executive sponsor: This person sets the goals and aligns

customer stakeholders. They figure out who owns what

CHAPTER 4 Ten Generative AI Success Factors 35

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 responsibilities, get the team going on implementation, clear
roadblocks, and make sure everyone knows why the project
is vital. The executive sponsor also keeps the budget flowing.
»» Program owner: This person is especially hands-on in
running the show, coordinating teams, prioritizing use cases,
mapping workflows, and planning the roll-out. This role
ensures that end-users have the resources and training to
know what they’re doing.
»» Admins and team leads: These roles help identify use cases
with their subject matter expertise and incorporate AI into
workflows. They’ll set up guidelines and templates, suggest
training examples, and in general, serve as power users.
»» Technical managers and IT contacts: These folks handle
technical nuts and bolts such as security and access, logins,
integrations, installation, and testing.
»» End-users: This group is everyone else on the front lines —
the people using the new AI capabilities in their daily
workflows. Their jobs are to train themselves, share feed-
back, and create success stories that you can share.

Spreading Impactful Use Cases

Chapter 2 and 3 provide examples of use cases for generative
AI across multiple teams and functions of your organization. I
share real-world examples and offer insights into how your orga-
nization should map out and prioritize use cases for your own
implementation.

Now comes the time to make it happen on a broad scale. Your use
case mapping helped you identify places to start. Keep in close
touch with end-users because you want to harvest their success
stories to help with the buy-in elsewhere. It’s a safe bet that once
other teams start to hear about what generative AI can achieve,
they’ll be lining up to operationalize it in their areas, too. That’s
part of the change management addressed in the next section.

36 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Putting People First in Change
Management
Like any big evolution, implementing your generative AI program
requires careful change management planning. You need effective
internal communications to talk up advantages and calm fears.
Change champions can also help facilitate the transition and
answer the “what’s in it for me” questions.

An early adopter program can work with those most ready for
change in order to generate early wins. Your team leads are
among the power users who are adopting early and spreading
both expertise and enthusiasm.

Change management is a cross-functional endeavor. A change

governance framework helps your enterprise codify a successful
path toward AI adoption. It involves all stakeholders, manage-
ment and frontline employees, and external experts.

Careful change management includes establishing a corporate

governance team and setting guidelines for how generative AI is
used and when it’s not. Carefully define roles and responsibilities,
monitor performance, and make any necessary adjustments.

Establishing Guardrails for Brand

Safety and Consistency
Generative AI can enhance your corporate reputation and promote
messaging that matches your organization’s values. But there are
steps to take to ensure success. For example, verify the unique-
ness and accuracy of generated content before using it in mar-
keting. Claim detection tools, such as those offered by Writer’s
generative AI technology, can help with your validation.

No matter how many powers you tap into and tasks you shift to
AI, the humans in the loop are essential. A successful AI imple-
mentation supercharges the creative capabilities of your people.
And it also relies on your people to establish the guardrails that
protect against bias and unethical uses.

CHAPTER 4 Ten Generative AI Success Factors 37

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 Transparency is essential. It ensures that your systems are fully
understood and their actions fully explained. It’s essential for
keeping your program accountable to the humans running the
show.

Setting Governing Principles

Responsible AI should be a cornerstone for any AI initiative, and
generative AI is no exception. Companies need to ensure that their
implementation adheres to an AI governance framework that
includes data privacy, fairness, explainability, accuracy, trans-
parency, interpretability, regulatory compliance, security, and
risk management. If your company doesn’t have an AI governance
board, consider starting one.

As your organization adopts AI, keep the essential principles for

AI adoption in mind, such as human-centric focus, clear purpose,
respect for safety, respect for human autonomy, privacy protec-
tion, respect for human dignity, data transparency, auditability,
and security. Data security and privacy always should be top pri-
orities included in your generative AI implementation. Know how
your data is being used in training the language model.

Becoming an AI-Ready Organization

Tapping into the capabilities of generative AI is a tantalizing
prospect. Is your organization ready? Here’s a list of questions to
ask yourself, so grab a piece of paper to document your answers:

»» Have you identified the top use cases for generative AI? Try
writing out your top five examples — and give yourself extra
points for identifying the key inputs and ideal outputs.
»» Do you have IT and executive buy-in for the initiative? Write
down the names of the people who still need to be brought
onboard, and think about how to get their buy-in.
»» Have you established a set of key performance indicators
(KPIs) for the generative AI proof-of-concept that you will
start with? Jot them down.
»» Do you know how you’ll calculate the return on investment
(ROI) of your generative AI investment? Spell out your metrics.

38 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 »» Have you established a project team to implement the
generative AI proof-of-concept? Make a list of who’s on
the team, and include their roles.
»» Do you have sensitive data that you want to maintain
ownership of? If so, ponder out loud how you intend to keep
this data secure/private in your generative AI solution.
»» Do you know how to integrate proprietary or internal
knowledge data sources securely into generative AI output?
»» Have you identified the line-of-business applications that will
be needed to support the proof-of-concept? What are they?
»» Do you have a set of customized AI guardrails that need to
be implemented? Scribble out some examples.
»» Do you have a plan for administration, operating, and
reporting on the generative AI platform? Think about who’s
in charge and ways they’ll report on how it it’s going.

Carefully Evaluating Vendors

You have multiple options as you seek a vendor partner for your
AI program. Be sure all your needs are fully addressed and that
you ask all the pertinent questions. Check out the end of Chapter 3
for more about these kinds of questions.

In broad terms, you want details about the technology and what
kind of LLM is used. There are important questions about data
sources and management. Customization and integration are
key considerations, and you have a lot of queries about enter-
prise security as well as legal and regulatory compliance. Ask good
questions about bias prevention, how solutions handle scalabil-
ity, and what kinds of monitoring and reporting are available. Of
course, you’ll have plenty of questions about costs, too.

Building Ongoing Training

“If you build it, they will come,” according to a well-known
movie line. But if you build generative AI into your workflows,
will your employees know what to do with it? That’s where train-
ing becomes critical.

CHAPTER 4 Ten Generative AI Success Factors 39

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 It begins with a company-wide AI use policy designed to help
employees understand what the aim is and avoid common pitfalls
and security risks. And some team members may be able to get a
head start by making better use of the generative AI through tools
they already use. Indeed, your adoption will be much smoother if
you choose an AI platform that integrates enterprise-grade gener-
ative AI functionality right into your commonly used applications.

From there, it’s a matter of building muscle memory, helping

employees know when and how to employ AI technology and
weave it into their workflows. They also need guidance and prac-
tice in writing the right prompt, which makes all the difference in
getting good results.

Onsite training, workshops, and webinars will be essential. The

right vendor can help you navigate the complex landscape of AI
technologies and help you select the right ones, and help you get
your team up and running.

Shifting the Hiring Mindset

AI is not about eliminating jobs, but it definitely will change jobs,
and in the longer-term, that may evolve how your HR team looks
at the skillsets of candidates. Generative AI will let you strategi-
cally redesign roles as it makes room for more value-added work
by humans.

There are some things to think about as you evolve. For one thing,
while automating menial tasks will give you a productivity boost,
if you end up with fewer entry-level roles, that means fewer
opportunities for on-the-job training and ground-up promo-
tions. Consider creating intentional, apprenticeship-style pro-
grams to attract and cultivate the new and talent you’ll need later.

Inevitably, AI will increase the need for experienced judgment and

subject matter expertise. You may want to hire an AI program
director to guide your adoption of AI technology and make sure
your business is prepared to acquire and develop a successful, AI-
informed, workforce.

40 Enterprise Generative AI For Dummies, Writer Special Edition

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
 And for those already on the team, they’ll need help determining
how to use AI to expand the reach of their day-to-day work. Equi-
table upskilling will help make your workforce more adaptable.

Thinking Big
Don’t diminish your possible success by failing to fully use your
imagination. It’s important to think big, and when you do, you
can increase the output of your creative teams by an astonishing
multiple.

It takes the right platform to achieve the most success. But

ultimately, although I’m talking about technology, your success
still derives from the people who are employing it. Succeeding
with generative AI involves collaborating across team functions
to discover and innovate, committing to act responsibly, and edu-
cating your people.

And one more set of people to mention: the ones working for
the vendor you choose. They help you tap into out-of-the-box
solutions when possible and work with you to develop tailored
solutions that integrate with your workflows.

CHAPTER 4 Ten Generative AI Success Factors 41

These materials are © 2024 John Wiley & Sons, Inc. Any dissemination, distribution, or unauthorized use is strictly prohibited.
WILEY END USER LICENSE AGREEMENT
Go to www.wiley.com/go/eula to access Wiley’s ebook EULA.