19,99 €
Mehr erfahren.
- Herausgeber: John Wiley & Sons
- Kategorie: Fachliteratur
- Sprache: Englisch
Cut through the noise and unlock the transformational power of conversational AI
In the newly revised second edition of Age of Invisible Machines, renowned tech leader Robb Wilson delivers a startlingly insightful and eye-opening blueprint for using conversational AI to make your company self-driving—with a digital ecosystem of interconnected automations powering all aspects of your business.
Conversational AI is transforming every job at every company (starting yesterday) and this book is perfect for anyone affected by these technologies. You'll learn how to develop a hyperautomation strategy by identifying outdated processes and systems holding your company back.
This latest edition offers brand new chapters dedicated to fast-growing automation tools, including Large Language Models, generative AI, and much more. You'll discover ways to implement new technologies that are force-multipliers for rapid growth. A must-read for every business leader, Wilson's book debunks common myths about conversational AI while simplifying the inevitable complexity of restructuring your business to unlock the substantial opportunities this new era offers.
You'll also find:
- Incisive discussions of the ethical dilemmas that lie before us as mass adoption of conversational AI takes effect
- Fascinating examinations of what a self-driving business looks like and how you can use conversational AI to generate an enduring competitive advantage
- Strategies for creating a hyperautomated ecosystem that any company can begin using immediately
- QR links to interactive and ongoing discussions of the subjects covered in each chapter
A practical and essential exploration of the future of conversational AI and hyperautomation, Age of Invisible Machines belongs in the hands of entrepreneurs, founders, business leaders, tech enthusiasts, designers and anyone else with a stake in the future of business.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 493
Veröffentlichungsjahr: 2025
Ähnliche
Table of Contents
Cover
Table of Contents
Title Page
Copyright
Preface
Acknowledgments
Introduction
PART I: Imagining an Ecosystem of Orchestrated AI Agents
CHAPTER 1: Organizational Artificial General Intelligence Is Already Upon Us
Some Are Calling This Hyperautomation
OAGI, Baby Steps Toward Singularity
Let Orchestrated AI Agents Run Things
Hyperautomation Brings Hyperdisruption
Notes
CHAPTER 2: What Agentic AI Is—and Isn't
Dispelling Common Myths About Agentic Artificial Intelligence
Notes
CHAPTER 3: Competing in the Age of OAGI
What Successful Hyperautomation Can Look Like
Graphical User Interfaces Have Value, But Can't Scale
The Super UI of the Future
To Stay Competitive, Embrace Systemic Change
Notes
CHAPTER 4: The Ethics of Experiential AI
Notes
CHAPTER 5: How Organizational AGI Can Change the World
OAGI Swings a Mighty Axe
Where There's Hyperautomation, There's Hyperdisruption
Notes
CHAPTER 6: This Journey Has Been Personal
Note
CHAPTER 7: Learning the Terms
Notes
PART II: Planning an Ecosystem for Orchestrated AI Agents
CHAPTER 8: The Dream vs. Reality
Note
CHAPTER 9: Ecosystem Evolution Explained
The Four Evolution Phases of an IDW
Tracing an Example Through All Four Evolutionary Stages
Evolving Agents Toward OAGI
Notes
CHAPTER 10: Teams and the Co‐Creation Mindset
Meet the Team
A Day in the Life of a Strategic Liaison
Notes
CHAPTER 11: How to Architect Tools Like LLMs, Agents, and Generative AI
LLMs Alone Can't Turn Words into Action
Giving AI Agents Real Agency
This Is a Totally Different Approach to Software
Orchestrating LLMs, Agents, and Generative AI Requires an Open System
Feasibility Is All About Speed
With Speed Comes Momentum
Notes
CHAPTER 12: Organizational AGI Demands an Extensible Cognitive Architecture
Tools and Toolkits Are Locked in Boxes
Point Solutions Lack Flexibility
Extensible Cognitive Architecture Is the Answer
Prepare for Your Climb
Find Your Footing
No Code Means Fewer Barriers
Infrastructure Topology for Agentic AI
Microservices at the Core
Making the Case for an Uphill Climb
Notes
CHAPTER 13: Digital Twins in SPACE
Finding SPACE for Feature Reduction
Orbiting OAGI
Your Cobbler's Kids Should Have Moon Boots
Notes
PART III: Building an Ecosystem for Orchestrated AI Agents
CHAPTER 14: Orchestrating AI Agents to Become Organizational Operating Systems
Single‐Agent Systems Powered by LLMs
Multi‐Agent Systems Powered by LLMs
Digital Twin Creation
Metacognitive Capabilities
Advanced Problem Solving
Embracing Multimodal Environments
Harnessing Collective Intelligence
Swimming with Orchestration
Nothing If Not Agile
Create an Enterprise Road Map
Off the Markov
Let Journey Maps Lead the Way
Challenges and Limitations
Notes
CHAPTER 15: Design Strategy for Organizational AGI
Design for Human‐Controlled Outcomes
Sequencing Patterns for a Successful Ecosystem
Key Patterns to Sequence for Conversational AI
Patterns in Action
Advanced Design Patterns for Agentic Orchestration
Notes
CHAPTER 16: Production Design for Organizational AGI
Everything Hinges on Analytics and Reporting
A Fast and Fluid Feedback Loop
Strive for Adaptive Design
Make Your Data Consumable
Best Practices in Analytics and Reporting
Note
CHAPTER 17: Best Practices in Conversational Design
Two Important Morphisms
PART IV: Conclusion
CHAPTER 18: Where Do We Go from Here?
Companies Running Scared from Consumers
A New Relationship to Work
Healthier Citizens and Healthier Governments Forging Balance Amid Chaos
Software Will Always Be About Communication
We Still Just Want to Be Happy
Notes
About the Authors
Index
End User License Agreement
List of Illustrations
Chapter 1
FIGURE 1.1 An ecosystem of intelligent digital workers.
Chapter 3
FIGURE 3.1 When companies find their stride with hyperautomation, it becomes...
FIGURE 3.2 Hyperautomation efforts without no‐code and conversational techno...
FIGURE 3.3 Hyperautomation with no‐code and conversational technologies.
Chapter 5
FIGURE 5.1 David Latimer's terrarium. (www.solentnews.co.uk)
Chapter 7
FIGURE 7.1 Example of a skill from the OneReach.ai GSX platform.
FIGURE 7.2 A conversational designer sequences steps in their flow in the On...
FIGURE 7.3 A human agent uses OneReach.ai GSX HiTL tools.
Chapter 9
FIGURE 9.1 Four evolutionary phases that an IDW can move through.
FIGURE 9.2 The data and information literacy stage.
FIGURE 9.3 The knowledge stage.
FIGURE 9.4 The intelligence stage.
FIGURE 9.5 The wisdom stage.
FIGURE 9.6 The data/information stage through the wisdom stage.
Chapter 10
FIGURE 10.1 Meet the team.
Chapter 11
FIGURE 11.1 Three‐quarter view of an IDW and its skills with height indicati...
Chapter 12
FIGURE 12.1 The iron triangle, or triple constraint.
FIGURE 12.2 The flexibility/usability trade off.
FIGURE 12.3 Traditional automation architecture.
FIGURE 12.4 End‐to‐end architecture.
Chapter 14
FIGURE 14.1 Some of the many functions within organizations where skills are...
FIGURE 14.2 Typical application of a Markov chain.
FIGURE 14.3 Attempting to map multiple conversations using a Markov chain ge...
FIGURE 14.4 Co‐creation process at a glance.
FIGURE 14.5 Explore vs. exploit.
Chapter 15
FIGURE 15.1 Is this porcupine approachable?
FIGURE 15.2 Key design patterns to sequence for conversational design.
FIGURE 15.3 Key design patterns to sequence for conversational design.
FIGURE 15.4 Advanced design patterns for agentic orchestration.
Chapter 17
FIGURE 17.1 Skeuomorphic design as it relates to conversational design. (One...
Guide
Cover
Table of Contents
Title Page
Copyright
Preface
Acknowledgments
Introduction
Begin Reading
About the Authors
Index
End User License Agreement
Pages
iii
iv
vii
viii
ix
x
xi
xii
xiii
xiv
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
53
54
55
56
57
58
59
60
61
62
63
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
296
95
96
97
98
99
100
101
103
109
104
105
106
107
108
110
111
112
113
115
116
117
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
197
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
249
250
251
252
253
254
255
256
257
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
279
280
281
282
283
284
285
286
287
288
289
290
291
292
301
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
REVISED AND UPDATED SECOND EDITION
Age of Invisible Machines
A guide to orchestrating Al agents and making organizations more self‐driving
Robb Wilson
with Josh Tyson
Copyright © 2025 by Robb Wilson. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.
Published by John Wiley & Sons, Inc., Hoboken, New Jersey.
Published simultaneously in Canada.
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 Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per‐copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750‐8400, fax (978) 750‐4470, or on the web at www.copyright.com. 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/permission.
Trademarks: Wiley and the Wiley logo 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. 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: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.
For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762‐2974, outside the United States at (317) 572‐3993 or fax (317) 572‐4002.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our website at www.wiley.com.
Library of Congress Cataloging‐in‐Publication Data is Available:
ISBN: 9781394321551 (Cloth)
ISBN: 9781394321568 (ePub)
ISBN: 9781394321575 (ePDF)
Cover Design and Image: by north.no & Daryna Moskovchuk
Author Photos: Courtesy of Mandraketheblack.de
Preface
By Josh Tyson
Much has changed since the first edition of Age of Invisible Machines was released back in the fall of 2022. I give Robb credit for astutely predicting the incoming adoption of conversational technologies, as ChatGPT was unleashed on a largely unsuspecting world a couple of months after our book was published. We both watched with excitement and awe as OpenAI’s public‐facing large language model (LLM) sent shockwaves through every industry and prompted more and more individuals to take a deeper look at the technologies associated with conversational AI, and decide how they might use them in their own lives.
Robb and I also watched with excitement and awe as Age of Invisible Machines quickly became the first bestseller about conversational AI. That unexpected success has afforded us the opportunity to put together this second edition in paperback. While the core information in the first edition remains practical and strategically sound, there have been many radical new developments that we were eager to put into context. Without fully realizing it, we were preparing for the task when we launched the Invisible Machines podcast shortly after the first edition came out.
We wrote the first edition of this book over the course of many hours of conversation, and we decided to hit record and continue our investigations, inviting in the smartest people we could find who are working with, or thinking deeply about, these technologies. You’ll see that insights and commentary from episodes of Invisible Machines have made their way into the pages of this second edition. There’s also a new kind of digital experience connected with this tangible one you’re holding in your hands that can connect you with content from the podcast, as well as the many other resources we reference throughout.
Using the practical information we’ve been sharing (and with Robb’s help) I’ve done some AI agent orchestration of my own. Follow any of the QR codes in these pages and you’ll begin a journey with an intelligent digital worker (or IDW) that has a nuanced understanding of the ideas we talk about in this book. Think of it as a revolutionary sidecar experience: as you read, you can have a conversation with Invisible Machines (literally and figuratively).
Along with updates throughout the text, we’ve completely rewritten chapters 11, 12, and 13 to better align with the current marketplace. And while we’ve been told this book is a pretty good read straight through, you can get something useful from reading any one of the chapters herein. One last note, most of the chapters are written from a “we” perspective to mirror the tone of the Invisible Machines podcast that Robb and I co‐host. There are other chapters, however, where Robb is sharing his direct perspective on his decades‐long journey with conversational AI, so the narrative shifts to his “I” perspective.
With that, welcome to the exciting, expanding universe of Invisible Machines!
Follow the QR codes found at the end of each chapter to continue your learning journey with an intelligent digital worker (or IDW) that has a nuanced understanding of the ideas expressed throughout this book as well as insights uncovered on the Invisible Machines podcast.
Acknowledgments
I've been collecting the many ideas, experiences, and stories that went into this book for decades, and getting them wrestled out of my head and onto bookshelves fell on the capable shoulders of Elias Parker, this project's producer and developmental editor. Many thanks for your commitment to seeing this project through. Thank you also to Josh Tyson for helping give these many ideas, experiences, and stories structure and a voice on paper.
Hearty high‐fives to Alison Harshberger, Alla Slesarenko, Vira Prykhodko, Mykhailo Lytvynov, Daryna Moskovchuk, Cole Gentile, Katie Tymchenko, Michael Salamon, Henry Comes‐Pritchett, Julie Kerr, the wonderful team at Wiley, and The Editrice, Kirsten Janene‐Nelson. Additional thanks to Melody Ossola for your visual designs and bynorth.no for your cover designs. My journey in conversational AI has brought me into contact with so many great people, incredible opportunities, and sizable challenges, and Age of Invisible Machines isn't the product of just one person, it's the sum of many parts.
To that end, I'd like to thank the many people who have joined the ongoing conversation on the Invisible Machines podcast. We've been fortunate to welcome many of the world's brightest minds into the fold and the ideas they shared were instrumental in shaping the revised version of this book. Special thanks to Cassie Kozyrkov, Jeff McMillan, Ben Goertzel, Cathy Pearl, Annie Harshberger, Blaise Agüera y Arcas, Rebecca Evanhoe, James Bridle, Ovetta Sampson, Richard Saul Wurman, Tom Gruber, Adam Cheyer, Laura Herman, Tim Wood, Lou Rosenfeld, Dr. Lee Hood, Dr. Nathan Price, Don Scheibenreif, Greg Vert, Don Norman, Seth Godin, and Professor Daniel Lametti.
A special thank you to my business partners, Daisy and Rich Weborg, and Kevin Fredrick. Huge thanks all around to Michael Bevz, Lance Christmann, Jonathan Anderson, Petro Tarasenko, Helen and Antony Peklo, Natalia and Andrey Nikitenko, and the entire family of amazing people at OneReach.ai (who've been diligently building the best AI agent orchestration platform in the world for many, many years now). Thank you to our customers and partners for recognizing the power of our approach long before Gartner, Forrester and others came along—especially Sherry Comes, former IBM Watson Distinguished Engineer and CTO, and Jordan Ratner, Senior Generative AI Strategist at AWS. Thank you also to the old Effective UI crew and the community of authors and readers we've built up at UX Magazine over the years.
My eternal gratitude goes to the powerful women I've known throughout my life, including my beautiful daughters: Sid, Cole, Katie, Melly, and Quinn. Opa (“Billing”), you rock. Mom, thank you for everything. Sasha, thank you for doing this with me. I love you. Thank you to my amazing siblings: Holly, Burt, and Ernie. Finally, thank you Marshall McLuhan, for setting my evolving worldview into motion so many years ago.
—Robb Wilson
Along with Robb and the many wonderful people mentioned above, I'd also like to thank my mom for everything. Eternal gratitude to Elias, Arius, Obie, Scout, and my amazing partner in life, Nicole. LYF!
—Josh Tyson
Introduction
Agentic AI, My White Whale
By Robb Wilson
Call me Ishmael.
Actually, please don't.
Like Captain Ahab in Moby Dick, I've spent many a waking hour in the heated pursuit of a powerful and elusive white whale: agentic AI. For lingering days, months, and years I've chased this steely beast on the horizon. I was frequently knocked off course by the complexity and newness of the various associated technologies, but I kept up the chase on both sides of land and over all sides of earth (as Melville would say).
Several years ago now, Josh Tyson and I started the surprisingly conversational experience of turning my journey into a practical guide for organizations that were ready to take the plunge early on in an inevitable shift in our relationship with technology. When the book quickly became a business bestseller, we created the Invisible Machines podcast to continue our conversations about these explosive technologies with the smartest people we could think of (including Cassie Kozyrkov, Dan Goldin, Ovetta Sampson, Richard Saul Wurman, Ben Goertzel, Laura Herman, Kara Swisher, Lisa Feldman Barrett, Jaron Lanier, Paul English, Cathy Pearl, Jim Webber, Seth Godin, Tim O'Reily, Adam Cheyer, Tom Gruber, and Tim Wood). So while this book is about a journey in technology, I don't think of it as mine alone.
When ChatGPT revealed to the general public just how revolutionary the act of conversing with a machine could be, Josh and I expected businesses to do some self‐reflection and make moves toward adopting AI in earnest. Instead, there's been a fair amount of hand wringing by organizations while tech giants looked for the fastest ways to bolt large language models (LLMs) onto existing products. It's been disheartening, but that's not to say that there aren't forward‐thinking leaders, designers, researchers, developers, and product designers pushing their companies in the right direction. We've had the good fortune to work with many organizations that are taking this moment seriously.
The biggest surprise has been watching everyday users out in the world really putting this technology to the test. In a way, consumers are leading the charge, and I'm starting to wonder if rather than AI in the hands of companies eliminating jobs, it might be AI in the hands of people eliminating companies. That might sound drastic, but it would be easy for customers to express their dissatisfaction with a company by orchestrating some AI agents to totally overrun their call center. At present that wouldn't even be an illegal action. Those aren't the kinds of applications we're talking about in this book, but it's getting easier to imagine consumer‐led activities sending disruptive ripples through any industry.
Organizations that see the bigger picture are already finding ways to make their operations more nimble, so that they can meet new challenges across unpredictable terrain. There are golden opportunities for people to collaborate in designing a future where technology is a partner that's helping us forge a more balanced and equitable world for all humans. AI won't erase our many mounting problems, but we can let it help us make more efficient decisions in responding to the fallout from something as unpredictable and complex as climate change.
My journey in conversational AI began as an early practitioner in the field of experience design. I noticed that the absolute worst experiences people were routinely having with machines were conversational in nature: purgatorial voice‐automated call centers and feeble chatbots trying to solve problems online and wreaking havoc on user trust with their inefficiency. Lifting users and organizations out of the seemingly infinite shitbot doldrums seemed like the true calling of experience design.
It seemed like nothing could make interactions with machines easier than using our most natural forms of communication—effectively masking the machinations, making the machine invisible.
As you'll discover in these pages, not only can agentic AI running behind the scenes in an organization handily obscure the mess of systems (and graphical UIs), it also binds your ecosystem by standardizing communications, creating a feedback loop that can evolve automations for all your users—customers and employees.
In the course of building a platform for orchestrating AI agents, I've come to realize that success is really based on a combination of three key technologies: conversational user interfaces, composable architecture, and no‐code rapid application programming. In essence, this trinity is truly the white whale of experience design. It furthers the frictionless conversational interface to a point where software creation becomes democratized. Instead of focusing on the experience of using software, code‐free allows us to elevate our efforts to designing the conversational experience of creating software using composable architecture. Not only do these component technologies elevate UX design to new heights, they can do the same for humanity, ensuring everyone has access to technology and can easily put it to use: technology that doesn't leave anyone behind.
The requirements for this level of automation—hyperautomation—create a scenario where the various elements of experience design become part of a rapidly moving feedback loop. For lack of a more creative term, this represents a kind of hyperUX. The energy previously poured into designing around the limitations of graphic UIs now gets poured into designing for an infinitely scalable conversational interface. The architecture of these experiences requires extensive journey mapping, and those maps become living documents that evolve alongside the experiences. Extensive research and analytics can happen in real time as the people inside your organization watch users moving through experiences as they unfold. This creates an environment that's more agile than Agile, where code‐free interactions with machines allow for iterations and improvements being made constantly.
This might sound far‐fetched to anyone who's been held hostage by the sluggish development cycles of their key vendors, but in an ecosystem built for orchestrating AI agents, you can pull in any piece of technology and make significant changes to any aspect of a user's experience at a moment's notice. Taking the wrong approach to integrating the technologies associated with AI feels eerily like Ahab's fate: his ship split in two by a forceful beast he never properly reckoned with, he finds himself tied to that force and dragged out to sea. Getting it right involves recognizing the true nature of this swirling mass of tools and strategies, as something beautiful and monstrous and beguiling.
This endeavor doesn't amount to pursuing a trophy you win with a hundred harpoons. Those who try to conquer these technologies will find their schemes smashed to bits. The trick is to be nimble and swift, riding the waves alongside them as everything churns madly around you. Learn to harness and ride AI agents, composable architecture, and no‐code creation by creating a strategic environment where everyone in your org can use technology effectively and no one is left behind (or out to sea).
In the pages to follow, we'll take a deep dive into the complexity of this new realm. By sharing with you everything I've learned, I want to help you and your organization swim with the speed, strength, and flexibility necessary to propel itself forward in choppy waters. This will be an arduous and complex journey fraught with difficult decisions, but it will also give you the chance to reflect on the outdated processes and systems that have been holding you back. Regardless of whether this sounds appealing to you or horrifying, the bottom line is it's time to take the plunge. If you're going to drop into uncertain and icy waters, it's incredibly helpful to have a point of view, so I'll lend you mine.
PART IImagining an Ecosystem of Orchestrated AI Agents
Creating an ecosystem that supports the orchestration of AI agents is a monumental task, and you'll want to approach it with solid foundational knowledge. In Part I, we'll take a look at the current landscape (as of this writing), dispel some of the common myths associated with conversational technologies, and start to paint a picture of what successfully achieving organizational artificial general intelligence (or OAGI) looks like. With technology this all‐encompassing and powerful, it's also crucial to bear in mind the myriad ethical concerns that will arise. These orchestrated technologies will upend our world in rather sudden ways, transforming established chunks of our daily routines. This work must be done with more than just good intentions; we must pay close attention to the shifting outcomes of those good intentions as things begin to accelerate. And accelerate they will. Despite the massive reverberations ChatGPT sent through the world, we are more or less at the starting point; from here we can set a trajectory that benefits everyone, leaving no one behind.
Massive opportunities are waiting for the organizations that take these efforts seriously, and that are willing to make the sacrifices of letting go of outdated systems and structures. There is a world within reach where everyone has access to technology—where soul‐sucking jobs are a thing of the past, and companies are increasingly self‐driving. In this world, people are free to work together on the most interesting, creative problems, and not only within the confines of large companies. The strategic orchestration of AI agents in pursuit of OAGI is just a means to that end—to that place where technology benefits everyone equally.
This is the book about how to get to that world. It's about how to get your team and company to organize and operate in ways that are highly conducive to achieving and maintaining a self‐driving state. A company that reaches OAGI is a bit like someone in a state of ketosis—having starved a body of carbohydrates to burn for energy so it starts burning fat for fuel instead. The artificial general intelligence at the heart of OAGI represents a system that has self‐awareness. A smaller version of the broader concept of AGI or singularity, OAGI refers to a system that knows enough to understand and contextualize everything that’s happening at any given moment inside and across an organization. OAGI means you've reorganized your organization's insides (likely starving it of outdated tools and processes) so that it can exist in a far more potent and efficient state. To whet this new appetite, we’ll also explore some scenarios of where hyperautomation can take your organization (and the world).
CHAPTER 1Organizational Artificial General Intelligence Is Already Upon Us
We don't want to scare you, but the revised edition of this book comes with the same ominous warning as the first edition: the status quo is a death sentence. Most modern organizations are still run using systems and strategies that will likely seem comically outdated a few short years from now. That's because the strategic orchestration of technologies described in this book, like AI agents, generative AI, deep learning, blockchain, and code‐free development tools, will do a lot more than disrupt the ways we're accustomed to dealing with technology. The strategic orchestration of these technologies will obliterate existing models.
Back in 2010, former Google CEO Eric Schmidt claimed that humans were creating as much data every two days as we had in our entire history up until 2003, pointing to “user‐generated content” as the main source.1 This was more than a decade before users started turning generative models loose, creating exponentially more content. While Schmidt's stat was contentious, there's no doubt that we're creating and capturing more data than ever.2 In many ways, however, this wealth of information represents a failure. To say it's being poorly leveraged is a bit of an understatement, but all of this is changing.
The technologies surrounding conversational artificial intelligence are heading toward a point of convergence that is already fundamentally altering our relationship with machines. We all experienced the early stages of these changes with the arrival of OpenAI's ChatGPT in the public sphere. The experiences customers and employees have with businesses are being reshaped by the hallmarks of this convergence—putting those massive stores of data into action in ways that have upended entire industries. In the first edition of this book, we said that this might sound hyperbolic given the substandard chatbot experiences endemic to much of the automation happening in the world. We now live in a world where most of the people you walk past on the street have engaged in conversations with machines, either by prompting any of the widely available large language models (LLMs) like ChatGPT or through interactions with companies leveraging LLMs in automated experiences.
These technologies are becoming more and more sophisticated, but that hasn't changed the fact that people don't like task‐switching between a whole host of different software solutions.
For example, let's say you log on to your home security system website to cancel service. Asking their chatbot a question drops you down a funnel of FAQ menus where you learn, five minutes in, that cancellations can't be handled online. When you call the accounts department, you're confronted by a series of voice automations that feels like another funnel drop, so you start stamping “0” hoping for a shortcut to a live agent. Crap experiences like this can feel less productive than just waiting on hold and hoping you can remember the security PIN you created five years ago.
The “I don't have time—I'll do it the long way” mindset is symptomatic of the lackluster conversational experiences users are accustomed to having with machines. But that's starting to change. One of the key elements of this convergence of technologies surrounding agentic AI involves intelligent and evolving ecosystems designed for accelerated automation powered by one of humankind's oldest adaptations: conversation. Make no mistake, agentic AI isn’t going anywhere. In fact, it’s going everywhere.
Groundbreaking as they've been, innovations such as Alexa and Google Home hardly qualify as conversational AI, or AI agents. Asking smart speakers to issue weather reports, set a timer, or play a song are very limited and immature applications of agentic AI, though they hint at its nascent power. Smart speakers have completely upended the speaker industry, to the point where it might become difficult to find a new speaker for sale that doesn't have built‐in conversational capabilities. But how powerful does a smart speaker become when it's not limited to the things that Siri or Alexa can do? What happens when you can ask your speaker to play “Mr. Roboto” by Styx and then follow up with another request: “I want to buy a copy of the book that Marc Maron mentioned during the intro to his podcast today. I don't remember the title but see if there's a copy available from Powell's before looking on Amazon.”
What will happen is, a few minutes later, a text message could appear on your phone with a link showing a hardcover copy of Camera Man by Dana Stevens available on Powells.com. By replying, “Yes. Please buy” via text, you'd be communicating with the same AI agent that you initially spoke to—an umbrella conversational interface that has become your primary interaction point with most of the technology in your life.
Apple is making moves to improve Siri, and Amazon has coupled Alexa with generative AI, allowing users to request chat sessions. This is a step closer to the scenario described above, but chatting with an LLM is a different experience than talking to a system that can automate any number of tasks. Once scenarios like this are possible, you won't think of technology in terms of different apps, because you'll rarely need to open and interact with an app. Domo arigato, Mr. Roboto.
Take this a step further and imagine how different the experience of work would be if employees could ask a smart speaker for help and instantly engage with a conversational operating system for their company that connected them to all the relevant departments and data needed to make their work less tedious and more impactful. This is the essence of organizational artificial general intelligence, or OAGI (it's kind of fun to pronounce it, oh‐ah‐gee).
We've spent a lot of time talking about this concept on the Invisible Machines podcast (with guests Ben Goertzel and Aaron De Smet of McKinsey) and believe it's a central concept for forward‐thinking businesses to understand. Give an organization a shared conversational interface that can communicate across channels and leverage data in various forms and organizational self‐awareness is likely to emerge. Really, the “organization” part is the context that a generally intelligent machine needs to do things the way a team or company wants them done. With OAGI, a technology ecosystem is in place that can perform all sorts of tasks without needing to involve humans unless necessary.
OAGI represents a systemic understanding of the inner workings of a brand. This awareness allows technology to operate as an extension of the brand for those inside the company as well as customers. Such a technology ecosystem would also be able to identify patterns within an organization and make predictions about all aspects of operations. Obviously, this kind of functionality goes far beyond what LLMs alone are capable of, and we will discuss numerous aspects of LLMs that bear sober consideration.
Human conversation is broader than the spoken word, as we have many ways of communicating our thoughts and needs. Humans frequently incorporate gestures, facial expressions, visual aids, and sounds in conversation. As such, conversational AI encompasses a full breadth of what we call “multi‐turn” or “multimodal” interactions. Because they are part of an interconnected ecosystem, these multimodal interactions can leverage those massive stores of data we're continually creating—unearthing massive opportunities for personalization and precision.
Having a text conversation with an invisible machine might include that machine showing you part of a video to illustrate a point. If you're asking it to analyze a spreadsheet or data, it can draw you a graph on the fly to help visualize data points. If the interaction is ongoing and you're about to start driving, the interface can move to voice command. Multimodal experiences mirror normal parts of conversations between people, and that sophistication enables humans to wield technological functions and capabilities using our most natural interface. These bite‐sized user interfaces, or micro UIs as we call them, are dialogue‐driven and, just like human conversations, they can include all kinds of audio and visual aids and even haptic cues.
You could never have an experience this seamless and efficient while digging through nested tabs or apps—and many of the world's leading companies are coming around to this fact. Salesforce didn't just acquire Slack back in 2021. Their CEO openly admitted that they were rebuilding their entire organization around Slack, and that work continues as of this writing.3 They're betting that an integrated communication platform and a unifying conversational interface—one machine that connects to everything—will benefit customers, employees, and organizations in big ways. Slack has added generative AI to the mix, leaning on LLMs to summarize activity on users' channels. Microsoft's Copilot offers similar capabilities, but again, as of this writing, these amount to bolt‐on applications that fall short of the advanced orchestration of technologies detailed in these pages.
It's critical to establish an understanding of the difference between bolt‐on applications of AI technologies and real systemic change. Bolting things like LLMs onto existing systems can yield short‐term results that can easily fool those who haven't unpacked the concept of OAGI. Copilot users might save a little bit of time using generative tools to write emails, but many of those same people would save just as much time using publicly available tools, like ChatGPT. The other issue is that a conversational interface that's limited by the amount of information and autonomy provided by a point solution can only do a limited amount of real work.
Systemic change means that a conversational interface becomes a portal into something much more vast and meaningful. It's not impossible for point solutions to become part of systemic change, but only if they aren't locked into partial solutions like Copilot, Salesforce, or Workday. AI agents can manipulate our current software through APIs, using both graphical user interfaces and command lines—interfaces originally designed for human interaction. Over time this approach allows point solutions to cohere and evolve into systemic changes.
This transition from point solution to systemic change won't be abrupt but will occur progressively, bringing significant value at each stage. The organizations that make the leap first will likely build an insurmountable lead. This requires careful consideration, however, as many point solutions are not flexible enough to become part of systemic change and will lead to technology dead ends.
We'll dig deeper into the relationship between point solutions and systemic change in Chapter 12, but it should become clear quickly that having agentic AI operate our old software makes sense. Continuing to invest in the antiquated wares of software providers that are either threatened by systemic change or unaware of the inevitable is a major dead end.
The systemic change described in this book utilizes conversational AI to let users leverage natural modes of communication and engage in experiences that can orchestrate all of an organization's existing software and data. Any feasible strategy for integrating agentic AI into an organization has to hinge on systemic change. Beware: anything less amounts to random acts of automation that won't lead anywhere substantial.
Most Companies Are Dying Young
According to Yale professor Richard Foster, the average lifespan of an S&P 500 company has decreased by more than 50 years in the last century. As he told the BBC in 2018, their life span had dropped from 67 years in the 1920s to just 15 years, describing the rate of change as moving at “a faster pace than ever.”4
In an article from around the same time in Forbes titled “Why Some Companies Die Young,” Jeff Stibel makes an astute point that relates to the pursuit of OAGI. “What is missing in business—what everyone is missing—is that the unit of measure for progress isn't size, it's time. Business owners need to start thinking more about survival than size. Bigger isn't better. And growth for growth's sake is downright dangerous.”5
With those things in mind, we started thinking of this as a longevity book for companies, analogous to something like Peter Attia's Outlive: The Science and Art of Longevity. The practical strategies we've outlined for integrating the technologies associated with agentic AI have the key benefit of making organizations far more flexible and futureproof.
In the BBC article mentioned above, author Kim Gittleson points to an emphasis on innovation and reinvention (i.e., flexibility) as a heavy factor in the longevity of companies, reminding us that Nokia was a pulp manufacturer before it moved into electricity and, eventually, mobile phones.
Gittleson notes that Japan is home to more than 20,000 companies that are more than 100 years old (as well as a few that are aged more than 1,000). There's a specific word that describes these long‐lived companies in Japanese: shinise.
“Professor Makoto Kanda, who has studied shinise for decades, says that Japanese companies can survive for so long because they are small, mostly family‐run, and because they focus on a central belief or credo that is not tied solely to making a profit,” Gittleson writes.
The organizations that thrive during the age of invisible machines are likely to focus on a central belief that technology can move them beyond the experiences that humans alone can provide. What makes this belief compelling to the people inside these organizations is that they are the ones directly benefiting from the implementation of OAGI. A key component of our strategy involves putting employees at the center of innovation—ahead of customers and shareholders.
While this strategy won't make orgs family‐run, it does borrow from a prominent feature of a family‐run business: everyone has a stake in its success. Using technology to make better experiences for employees makes a business a better place to work. The work of creating these experiences moves you closer to OAGI.
It's impossible to ignore the fact that this level of automation will likely lead to fewer people working inside of organizations, making them smaller and perhaps boosting their longevity. That said, the goal isn't to replace people inside of orgs; it's to augment their skills and make their experiences better.
As such, these strategies require us to think about organizations differently. The good news for companies is that they no longer need to be confined by rigid software (and the exhausting process of building it). Orgs can stop retrofitting their work processes and start taking a new, more dynamic shape capable of making it in the long run.
As natural conversation becomes the primary interface between machines and the humans using them, the machine becomes invisible as the interface disappears. This line of thinking should be familiar to most experience design practitioners. One of the hallmarks of successful experience design is an interface that gets out of the way. The further the interface recedes into the background during an experience, the more frictionless that experience becomes. This lightens a user's cognitive load and helps them to get what they need from the technology more effectively (though it also represents a massive amount of orchestration behind the scenes).
With agentic AI, interfacing with machines no longer requires that we adapt to the way they communicate, which dramatically reduces friction in our experiences with machines and software. As we’ve said, agentic AI will go anywhere and everywhere—meaning that invisible machines will be, for lack of a less grandiose term, omnipresent. You'll be able to turn to your phone, any nearby smart speaker, or any voice‐enabled appliance and enlist the help of an invisible machine. This ties into another element of this convergence, which involves sequencing technology so that it can react and adapt to individual situations. Invisible machines galore, connected to ecosystems built for optimized problem solving. Using agentic AI to operate our existing software for us moves away from brittle approaches like robotic process automation (RPA). OAGI puts organizations in a position to write their own software that's far more agile and customizable than the wares of third‐party vendors—software that doesn't require scripts or specific instructions to complete objectives.
Some Are Calling This Hyperautomation
Sequencing disruptive, advanced technologies to work in concert is something Gartner calls hyperautomation, and it's as intense as it sounds. Gartner coined the term in 2019; by their estimation, “Technology is now on the cusp of moving beyond augmentation that replaces a human capability and into augmentation that creates superhuman capabilities.” We find Gartner's definition to be a bit loose. If you tighten the definition to specifically require that automation results in producing beyond human experiences, then there's less confusion around how hyperautomation is different from automation. Presume hyperautomation to result in better experiences, not just automated experiences, and that's how we'll be using the term in this book. For the most part it's also analogous to OAGI, though you could argue that OAGI is the direct result of being in a state of hyperautomation. As with broader AGI, there's also the idea of some kind of awareness, which would be a likely result from the activities surrounding hyperautomation. As an ecosystem is trained by humans to automate increasingly sophisticated tasks across an organization (i.e., hyperautomation), it develops deeper contextual awareness and can be trained to make all kinds of high‐value predictions. We'll make an effort to use the best term given the context of the information presented throughout these pages.
Implementing a strategy for hyperautomation (that will lead you toward OAGI) is a massive undertaking that requires cooperation from every department within your organization. But you're not helpless in the face of this massive change. You can't succeed by resigning yourself to this disruptive era of technology. It's not something you have to succumb to. You can take charge, but first you have to surrender many of the old ways of doing things.
Organizational AGI doesn't materialize because of concrete plans laid well in advance of action. Flexibility is the name of the game. Organizations need to forge a cognitive architecture that affords you control of the tools and software you use when automating experiences (something we will dig deeper on in Chapter 12). If you think of the broader goal being establishing OAGI, the idea of flexibility comes into focus. For humans, intelligence, or learning, is a lifelong process—an evolution. It's no different when it comes to OAGI.
Robb has been working on this stuff for decades, breaking down agentic AI into patterns that can be sequenced to automate real work. It's part of a deeper passion—for the majority of his life Robb has been trying to improve the ways people and machines communicate with each other. We've come to believe that the orchestration of AI agents will level the playing field. The many technologies associated with AI represent a massive leap forward on both sides of these interactions. These tools present the potential for disruption on a scale that will eclipse the advent of the printing press, the industrial revolution, and the dawn of the computer age itself. On top of that, the rate of change will be much faster.
Hyperautomation is really an application strategy that goes beyond the development of AI and into how you sequence it with other disruptive technologies to solve complex problems as part of an organization‐wide experience strategy. As you'll discover in Chapter 11, “How to Architect Tools Like LLMs, Agents, and Generative AI,” technology exists right now that can help you move toward hyperautomating business processes, workflows, conversations, and tasks to go beyond standard UX and into new territory. We think of it as going Beyond Human Experience (BHX). This means creating experiences that don't just use machines to replicate tasks the way humans currently perform them.
BHX is all about finding new ways to create experiences that surpass what humans alone are capable of. This doesn't mean that humans have to be in competition with machines. While there are certain roles that will undoubtedly disappear, ideally, machines will augment our abilities in ways that let them act as connective tissue between humans. Freed from tedium, we can extract more value from our interactions with others.
It's common for organizations to focus on disruptive technology in myopic ways—in this case, seeing automation as a means to handle simple tasks on human terms, such as automating a coffee maker so that it brews a fresh pot of coffee at 8:45 a.m. What if, instead, the coffee pot was part of a better‐than‐human experience that not only adjusts the time it brews coffee and the amount it brews, but also cross‐references company calendars and brews an extra‐strong pot of coffee in anticipation of a client coming straight to the office from an international flight.
To expand on this idea from the first edition of this book, the AI agent that manages the experience of making the coffee might swarm with other agents to automate the purchasing of the coffee—continually looking for the best price, following up to see if people like the coffee, and creating a feedback loop to hone a better coffee experience for everyone over time. This agent can obsess over the coffee experience, getting as granular as the amount of caffeine in different offerings.
This brings up the concept of AI agents as customers, something we'll explore further later on. We discussed the subject on the Invisible Machines podcast episode with Don Scheibenreif, a Distinguished VP Analyst at Gartner and co‐author of When Machines Become Customers. His book says that CEOs believe up to 20% of their companies' revenue will come from machine customers by 2030. Companies will need to rewire themselves to appeal to AI agents as buyers. Among many other adjustments, traditional marketing hyperbole will be a waste of money and time. Genuine user reviews will be critical. Truth will become nearly impossible to hide. This is systemic change.6
OAGI, Baby Steps Toward Singularity
When we think about AI we often think about the notion of singularity—the hypothetical point in time when a powerful superintelligence will surge past all human intelligence. There's also the notion of machines gaining artificial general intelligence (AGI) and, thus, the ability to learn any intellectual task that humans can. These versions of superintelligence won't be the product of some super algorithm.
The perception of singularity or AGI is more likely to emerge from an ever‐evolving ecosystem of algorithms and technologies sequenced in intelligent ways to work in concert. This singularity likely would be made up of contributions from different pieces of software engineered all over the world. These systems will likely speak to each other using human language, making APIs a thing of the past. These ecosystems of algorithms and sequenced technologies are very similar to the ones we describe in this book. It seems likely that organizational AGI might represent the early pieces of a broader, decentralized AGI (sometimes called “singularity”). That is to say that, across industries, the organizations that reach a state of hyperautomation will set the tone for their respective industries while also interacting with other companies in a similarly advanced state. The choices business leaders make in these early moments will set a mighty trajectory.
While the arrival of singularity could be decades away, we've already quietly passed a significant milestone. Users are beginning to have experiences with AI agents that are far more rewarding than what their human counterparts can offer (BHX). We recently worked directly with an enterprise retailer in the United States that, over the course of a year, used the strategies described in the first edition of this book to reduce the number of calls to stores by 47% and achieve an NPS (or Net Promoter Score) of 65 on their automated customer service experience. The initial pilot in 2022 led to a $3 million increase in gross profit, with a projected annual increase of $80 million. That's just a glimmer of the potential buried in these explosive technologies.7
Let Orchestrated AI Agents Run Things
In the spring of 2024, Google put a spotlight on AI agents, prediction machines that can be put to work on all sorts of tasks. These are integral to the intelligent digital workers (IDWs) concept that we introduced in the first edition of our book. Unlike typical AI agents, however, IDWs operate within carefully designed ecosystems that can easily incorporate existing software and current best‐in‐market solutions. Put more succinctly, IDWs are agents that orchestrate other AI agents. We'll explore all sorts of scenarios (both favorable and unfavorable) in the “How Organizational AGI Can Change the World” chapter of this section. But for now, imagine this kind of reality playing out in another context: your router goes down. You place a call to your service provider and are guided through all the necessary system checks quickly and elegantly by a conversational app.
Or, better yet, their hyperautomated ecosystem detects that your router is down and their conversational app reaches out to you before you even notice that you're offline. This is the work of an IDW. More capable than a typical AI agent, the IDW is connected to this service provider's ecosystem, which is a network of interdependent technologies, processes, people, and, of course, other AI agents.
To simplify things in this second edition, rather than continually drawing the distinction between IDWs and AI agents we'll just communicate these ideas in terms of “AI agents” and “orchestrated AI agents.” IDWs are really a collection of AI agents working together to complete tasks and solve problems, driven by a conversational interface. Customers and employees can ask an IDW for help and it can orchestrate AI agents behind the scenes to deliver reliable, actionable information, or to complete tasks on the user's behalf.
The orchestration of AI agents from the example above reaches out to you to follow up on an alert from a fellow machine in the maintenance department that your location had lost coverage. These AI agents can swarm around the task, isolating and troubleshooting the issue by running background tasks while it speaks with you to verify your account and location. An AI agent could ask you to send a video of the blinking lights on your router that can be analyzed using computer vision. This AI agent swarm simultaneously looks internally at your connection status, assessing and course‐correcting in seconds. Your router is back up and running within five minutes, and you didn't have to wait on hold for a human operator, because there are unlimited AI agents at the ready—or because they called you, having detected the issue before you did. After experiencing this kind of BHX firsthand, you'll never want to go back.
There's also another element to consider that relates back to the notion of AI agents as customers. It's likely that consumers will come to rely on some form of personalized AI agent that can act as a buffer between consumers and, in this case, a utility provider. It's possible that all of this back and forth about your connection status happened without your knowledge, with the provider's AI agents communicating directly with your own AI agents. Your AI agents might request a credit on your bill and let you know about the situation after it's been resolved. The personal agents might even suggest switching providers and provide a quick summary of other options, and then also possess the ability to execute with your sign‐off.
As AI agents are sequenced with other technologies to contextualize massive amounts of data within an ecosystem that can give customers and employees access to elevated problem‐solving capabilities, the world as we know it will change fundamentally, as visualized in Figure 1.1.
In this book we'll explore the components of a robust ecosystem for hyperautomating business processes, workflows, tasks, and communications, along with what a strategy looks like for evolving these ecosystems.
“The biggest thing that pushed me to convert to Lemonade was the utterly charming AI chatbot,” Juliette van Winden wrote in a Medium post dedicated to their chatbot, Maya. “24/7, 365, day or night, Maya is there to answer any questions to guide the user through the sign‐up process. Unlike the drag of signing up with other providers, it took me a total of two minutes to walk through all the steps with Maya… . What intrigued me the most, is that it didn't feel like I was chatting with a bot. Maya is funny and charismatic—which made the exchange feel authentic.”8
The fact that a comparatively unintelligent conversational interface was able to launch Lemonade so far should give you inklings of how powerful these technologies really are. Remember, too, that the potential created by organizational AGI is so vast that the marketplace advantage can be staggering for companies with these ecosystems already in place. Making the most of this hyperdisruptive moment in history (and not being left behind) requires a holistic undertaking that touches on all aspects of your business. Random acts of technology—like deploying disparate machines that exist in isolation—will leave your workforce and customers underwhelmed, leading to low adoption rates and the likely removal of the offending tech. A fully integrated approach, however, can bring about a totally new paradigm of productivity with unprecedented potential.
FIGURE 1.1 An ecosystem of intelligent digital workers.
Even the initial steps take time. Jeff McMillan, chief analytics and data officer at Morgan Stanley, told us it took his team nine months to train GPT‐4 on more than 100,000 internal documents. This work began before the launch of ChatGPT, and his team had the advantage of working directly with people at OpenAI, including Sam Altman, himself. They created a personal assistant that the investment bank's advisors can chat with, tapping into a large portion of its collective knowledge. They've also laid the foundation for organizational AGI.
“Now you're talking about wiring it up to every system,” he said, with regard to creating the kinds of ecosystems and strategies contained in this book. “I don't know if that's 5 years or 3 years or 20 years, but what I'm confident of is, that is where this is going.”9
“Wow, This Sounds Really Hard”
Hyperautomation is indeed a momentous undertaking. The easiest way to get started is often to automate internally first; start small by automating individual tasks and skills, not entire jobs. Some of these early automations might seem underwhelming, but the simpler you make your starting point, the sooner you can test and iterate. The sooner you test and iterate, the sooner you can roll out an internal solution. You'll continue testing and iterating on that solution, using the momentum to find new skills to develop, test, iterate on, and deliver. You'll fumble often as you grow legs, but that's part of the process, too. In the realm of hyperautomation, we are more agile than Agile (hyperagile, in a sense). With the right tools and budding ecosystem, the iteration process becomes so speedy that failures are often quick rewards that point to better solutions. Because fixes and new solutions can be tested and deployed quickly and at will, your organization can build on wins and gain speed.
