The AI revolution is moving at a breakneck speed. Organizations are beginning to invest in innovative ways to monetize their data through the use of artificial intelligence. Businesses need to understand the reality of AI. To be successful, it is imperative that organizations understand that augmented intelligence is the secret to success. Augmented Intelligence: The Business Power of Human-Machine Collaboration is about the process of combining human and machine intelligence. This book provides business leaders and AI data experts with an understanding of the value of augmented intelligence and its ability to help win competitive markets.
This book focuses on the requirement to clearly manage the foundational data used for augmented intelligence. It focuses on the risks of improper data use and delves into the ethics and governance of data in the era of augmented intelligence. In this book, we explore the difference between weak augmentation that is based on automating well understood processes and strong augmentation that is designed to rethink business processes through the inclusion of data, AI and machine learning.
What experts are saying about Augmented Intelligence
"The book you are about to read is of great importance because we increasingly rely on machine learning and AI. Therefore, it is critical that we understand the ability to create an environment in which businesses can have the tools to understand data from a holistic perspective. What is imperative is to be able to make better decisions based on an understanding of the behavior and thinking of our customers so that we can take the best next action. This book provides a clear understanding of the impact of augmented intelligence on both society and business."--Tsvi Gal, Managing Director, Enterprise Technology and Services, Morgan Stanley
"Our mission has always been to help clients apply AI to better predict and shape future outcomes, empower higher value work, and automate how work gets done. I have always said, 'AI will not replace managers, but managers who use AI will replace managers who don't.' This book delves into the real value that AI promises, to augment existing human intelligence, and in the process, dispels some of the myths around AI and its intended purpose."--Rob Thomas, General Manager, Data and AI, IBM
Author(s): Judith Hurwitz; Henry Morris; Candace Sidner; Daniel Kirsch
Publisher: CRC Press
Year: 2020
Language: English
Pages: xxviii+141
Cover
Half Title
Title Page
Copyright Page
Endorsements
Dedications
Contents
Foreword
Preface
Why This Book? Why Now?
Why You Should Read This Book
What Is in This Book
About the Authors
Chapter 1: What Is Augmented Intelligence?
Introduction
Defining Augmented Intelligence
The Goal of Human–Machine Collaboration
How Augmented Intelligence Works in the Real World
Improving Traditional Applications with Machine Intelligence
Historical Perspective
The Three Principles of Augmented Intelligence
Explaining the Principles of Augmented Intelligence
Machine Intelligence Addresses Human Intelligence Limitations
Human Intelligence Should Provide Governance and Controls
Summary: How Augmented Intelligence and Artificial Intelligence Differ
Chapter 2: The Technology Infrastructure to Support Augmented Intelligence
Introduction
Beginning with Data Infrastructure
What a Difference the Cloud Makes
The Cloud Changes Everything
Big Data as Foundation
Understanding the Foundation of Big Data
Structured versus Unstructured Data
Machine Learning Techniques
Dealing with Constraints
Understanding Machine Learning
What Is Machine Learning?
Iterative Learning from Data
The Roles of Statistics and Data Mining in Machine Learning
Putting Machine Learning in Context
Approaches to Machine Learning
Supervised Learning
Unsupervised Learning
Reinforcement Learning
Neural Networks and Deep Learning
Evolving to Deep Learning
Preparing for Augmented Intelligence
Chapter 3: The Cycle of Data
Introduction
Knowledge Transfer
Personalization
Determining the Right Data for Building Models
The Phases of the Data Cycle
Data Acquisition
Identifying Data Already within the Organization
Reasons for Acquiring Additional Data
Data Preparation
Preparing Data for Machine Learning and AI
Data Exploration
Data Cleansing
Feature Engineering
Overfitting versus Underfitting
Overfitting versus Underfitting for a Model Predicting Housing Prices
From Model Development and Deployment Back to Data Acquisition and Preparation
Chapter 4: Building Models to Support Augmented Intelligence
Introduction
Explaining Machine Learning Models
Understanding the Role of ML Algorithms
Inspectable Algorithms
Black Box Algorithms
Supervised Algorithms
Creating a Gold Standard for Supervised Learning
K-Nearest Neighbors
Support Vector Machines
Unsupervised Algorithms
Understanding Reinforcement Learning and Neural Networks
The Value of Machine Learning Models
Summary
Chapter 5: Augmented Intelligence in a Business Process
Introduction
Defining the Business Process in Context with Augmented Intelligence
Weak Augmentation
Strong Augmentation
Strong Augmentation: Business Process Redesign
Augmented Intelligence in a Business Process about People
Strong Augmentation for Predictive Digital Marketing Campaign Management
Redefining Fashion Retailer Business Models with Augmented Intelligence
Business Model Changes at The Gap, Inc., Using Algorithmic Fashion Predictions
Another Fashion Retailing Business Model Using Algorithmic Predictions: Stitch Fix
Hybrid Augmentation
Summary
Chapter 6: Risks in Augmented Intelligence
Introduction
Providing Context and Understanding
The Human Factor
Understanding the Risks of a ML Model
The Importance of Digital Auditing
The Risks in Capturing More Data
Why It Is Hard to Manage Risk
Seven Key Risks
1. The Risk of Overfitting or Underfitting
2. Changing Business Processes Increases Risk
3. The Risk of Bias
4. The Risk of Over Relying on the Algorithm
5. The Risk of Lack of Explainability
6. The Risk of Revealing Confidential Information
7. The Risk of a Poorly Constructed Team
Summary
Chapter 7: The Criticality of Governance and Ethics in Augmented Intelligence
Introduction
Defining a Control Framework
Creating Your Augmented Intelligence Control Framework
Steps in Your AI Control Framework
Conducting a Risk Assessment
Creating Control Activities
Creating a Monitoring System
Data Privacy Controls
On an Organizational Approach to Controls
Summary
Chapter 8: The Business Case for Augmented Intelligence
Introduction
The Business Challenge
Taking Advantage of Disruption
Disrupting Business Models
Advantages of New Disruptive Models
Managing Complex Data
Creating a Hybrid Team
The Four Stages of Data Maturity
Building Business-Specific Solutions
Making Augmented Intelligence a Reality
How Augmented Intelligence Is Changing the Market
Summary
Chapter 9: Getting Started on Your Journey to Augmented Intelligence
Introduction
Defining the Business Problem
Establish a Data Culture
Moving Forward with the Foundation
Taking the First Steps
Selecting a Project that Can Be a Reference for Future Projects
Warning Signals
Summary
Chapter 10: Predicting the Future of Augmented Intelligence
Introduction
The Future of Governance and Compliance
Emergence of Different Jobs
Machines Will Learn to Train Humans
New Techniques for Identifying Bias in Data
Emerging Techniques for Understanding Unlabeled Data
Emerging Techniques for Training Data
Reinforcement Learning Will Gain Huge Momentum
New Algorithms Will Improve Accuracy
Distributed Data Models Will Protect Data
Explainability Will Become a Requirement
Linking Business Process to Machine Learning Models
Summary
References
Index
