The Fourth Industrial Revolution & 100 Years of AI (1950-2050)

By Alok Aggarwal

The Essential Guide to Demystifying Artificial Intelligence and Understanding How It Is Already Transforming the World Around Us

 

"A deeply researched and beautifully put together treatise on the most disruptive subject of our times."
– Ashok Belani, CTO & EVP, Schlumberger

 

"Very cool and timely!"
– Tom Leighton, MIT Professor and CEO/Co-Founder of Akamai

 

Today, Artificial Intelligence seems mystical and magical to most people. While this wave of change will touch every aspect of our daily lives, most of us still do not understand AI's capabilities, limitations, or history. In this book, Dr. Alok Aggarwal – one of the early innovators and developers in this field – sets out to demystify Artificial Intelligence by explaining the science and engineering in non-technical terms we can all understand.

 

With very little math and no software code, this book provides crucial information for the following, primary audiences:

• Product managers, program leaders, business leaders, consultants, and investment managers who may not need to understand the minute details of AI systems but should have sufficient knowledge to discuss with clients and technology teams, thereby improving business processes.
• Students, especially graduate students in science, technology, engineering, mathematics, analytics, business administration, financial engineering, and related disciplines.
• Aspiring entrepreneurs and decision-makers who will be ideally suited to exploit these inventions almost immediately.

This book explains numerous applications of AI that are already being used in other vital inventions of the current and the Fourth Industrial Revolution, including the Internet of Things (IoT), Blockchains, Metaverse, Robotics, Autonomous Vehicles, Three-Dimensional Printing, inventions related to predicting, mitigating, and adapting to rapid climate change, and innovations related to gene editing, protein folding, and personalized healthcare.

 

ABOUT THE AUTHOR:
Dr. Aggarwal is the founder, CEO, and Chief Data Scientist of Scry AI, which provides innovative AI-based products, solutions, and services to enterprises across the globe. He received his Ph. D. from Johns Hopkins University and worked at IBM's T. J. Watson Research Center.
 

• Language: English
• Publisher: Alok Aggarwal
• Release date: Feb 26, 2024
• ISBN: 9798224236237

Book preview

Acknowledgements

This book project started because I published four articles regarding Artificial Intelligence (AI) in March 2018. After reviewing them, my wife Sangeeta Aggarwal started nudging me to expand on the theme and write a comprehensive book, which I eventually started in January 2021. In fact, she continued prodding me until I finished this book. Hence, I am extremely grateful for all her nudging and prodding as well as for enduring me for the last 33 months. I also thank Adeeti Aggarwal (my daughter) for teaching me the basics of Neuroscience, especially how little human society knows about this subject.

I had written most of this book by September 2022. However, when I gave a few chapters to some friends for review, their consistent feedback was that these chapters were extremely technical and would not be grasped by the intended audience, which included business leaders and non-STEM graduate students. Fortunately, Simon Golden and Amol Aggarwal (my son) acted as book coaches, reviewed various versions of the manuscript, and gave me extremely valuable feedback that made the final version more comprehensible. Indeed, no words can do justice to my tremendous gratitude to both of them.

I am also grateful to Brij Masand, Sanford Roberts, Himanshu Shukla, and Shailejeya Shukla for reading all seventeen chapters and providing invaluable feedback, which improved this book’s readability even further. In addition, the following people provided enormous encouragement, precious feedback, and valuable insights: Srinivasan Bharadwaj, Joseph Bradley, Purav Desai, Alessio Garofalo, Piyush Gupta, Naveen Jain, Ujjal Kohli, Hema Krishnamurthy, Su Le, Sandeep Sacheti, and Moshe Vardi. Kshitij Suri created many diagrams and figures for this book and Mahima Manoj compiled much of the bibliography. Also, Weston and Jenny Lyon for Plug and Play Publishing made the editing and publishing process extremely easy and seamless. Undoubtedly, without their help, I may never have completed this book (at least in its current form), especially since the field of Artificial Intelligence has been evolving at a furious pace.

This book covers developments regarding the Fourth Industrial Revolution, AI systems, and Data Science. Although I have provided many hypotheses, opinions, and predictions, this book surveys key inventions that comprise the Fourth Industrial Revolution. Therefore, in the bibliography, I have provided around 900 references (until July 15, 2023) and regret if I have missed any. Furthermore, since the main topics covered constitute independent fields of study and since each chapter is worthy of a separate book, I apologize if I have oversimplified these topics or omitted any critical aspects.

Table of Contents

The Fourth Industrial Revolution & 100 Years of AI (1950-2050)

Acknowledgements

About the Cover

Introduction

Chapter 1 - Vital Characteristics of The Ongoing Fourth Industrial Revolution

1.1. Key Inventions of The Current Industrial Revolution

1.2. The Structure of Scientific Revolutions

1.3. The Structure of Industrial Revolutions and Three Key Characteristics

1.4. Technology Takes Substantial Time Before Becoming Pervasive

1.5. Hype May Be Irrational Exuberance But Can Be Advantageous

1.6. Once Adopted, Ramifications of Key Inventions Were More Extensive Than Anticipated

1.7. Winners and Losers, Jobs Lost and Jobs Gained

1.8. Role of Basic Science and Governments in Scientific and Industrial Revolutions

1.9. Discussion

Chapter 2 - Genesis of Artificial Intelligence and a Scientific Revolution: 1950-1979

2.1. The Pioneering Question

2.2. The Beginning of the Boom Phase

2.3. Emergence of Machine Learning and A Scientific Revolution

2.4. Statistical Machine Learning, Single and Multi-Layer Perceptrons

2.5. Other Subfields of AI are Born

2.6. Commercial Applications and Use Cases

2.7. The Bust Phase and the First AI Winter

2.8. Discussion

Chapter 3 - The Second AI Winter and Resurgence of AI Between 1980-2010

3.1. Hype And Bust of Expert Systems

3.2. Reasons for Substantial Growth in AI Systems

3.3. Emergence of Support Vector Machines

3.4. Revival and Expansion of Deep Learning Networks

3.5. Progress in Commercial Applications

3.6. Discussion

Chapter 4 - Domains in Which Artificial Intelligence is Rivaling Humans: 2011-2023

4.1. The Rise and Fall of IBM Watson

4.2. Deep Learning Networks (DLNs) Yield Remarkable Results

4.3. Emergence of Generative Adversarial Networks and Diffusion Models

4.4. Spectacular Rise of Gigantic Deep Learning Networks Called Transformers

4.5. Key Domains Where AI Systems Have Improved Markedly

4.6. Rapid Growth in Research Publications Related to AI

4.7. Discussion

Chapter 5 - The Internet of Things, Smart Cities, Data, and AI

5.1. High-Level Architecture Regarding IoT

5.2. Convergence of IoT and AI and Their Applications

5.3. Smart Cities, Green Cities

5.4. AI Systems and Smart Cities

5.5. Obstacles in IoT Adoption

5.6. Future Growth and Hype Regarding IoT

5.7. Discussion

Chapter 6 - Data and AI in Predicting, Mitigating, and Adapting to Climate Change

6.1. Climate Change and Its Predictions

6.2. Uses of AI in Predicting Climate Change

6.3. Uses of AI in Mitigating Climate Change

6.4. Exploiting AI in Adapting to Climate Change

6.5. Agriculture, Data, and AI

6.6. If Unchecked, AI May Accelerate Climate Change

6.7. Discussion

Chapter 7 - Applications of AI for Enhancing Blockchains

7.1. Advantages of a Decentralized System for Financial Transactions

7.2. Basics of the Blockchain Architecture

7.3. Smart Contracts and Two Significant Applications for Blockchains

7.4. Combining Blockchains and AI is Extremely Beneficial

7.5. Blockchain, IoT, and AI

7.6. Crypto Coins and Non-Fungible Tokens

7.7. Discussion

Chapter 8 - Extended Reality, Metaverse, Data, and AI

8.1. Technological Ingredients of Metaverse

8.2. Use of Artificial Intelligence in Metaverse

8.3. Digital Twins and AI

8.4. Applications of Metaverse and Embedded AI

8.5. Impediments to Creating Metaverse

8.6. Impudent Predictions Regarding the Growth of Metaverse

8.7. Discussion

Chapter 9 - Robotics, Driverless Vehicles, 3D Printing, and AI

9.1. Market Size for Robots and Their Categories

9.2. Key Application Areas for Robotics and Current Impediments

9.3. Large Amounts of Data and AI Are Improving Robotics

9.4. Autonomous Vehicle Driving – Recent Improvements and Current Limitations

9.5. Three-Dimensional Printing (Additive Manufacturing)

9.6. Robotics, 3D Printing, and Artificial Intelligence

9.7. Discussion

Chapter 10 - AI For Gene Editing, Protein Folding, New Materials, and Personalized Medicine

10.1. Gene Editing and CRISPR

10.2. Using AI to Improve Gene Editing

10.3 Applications of Gene Editing and Related Concerns

10.4. Using Artificial Intelligence for Solving the Protein Folding Problem

10.5. Discovering New Molecules for Producing Better Materials

10.6. Exploiting AI to Improve Drug Discovery and Design

10.7. Uses and Current Limitations of AI in Precision Medicine

10.8. Discussion

Chapter 11 - Limitations of Contemporary AI Systems and Their Consequences

11.1. Audacious Predictions, Ominous Sentiments, and Sci-Fi Laws

11.2. Contemporary AI Systems Are Beset with Many Limitations That Existed in the 1970s

11.3. DLNs are Brittle and Hallucinate

11.4. Machine Endearment and Malware Injections

11.5. Additional Limitations of Transformers and Large Language Models (LLMs)

11.6. Important Consequences of These Limitations

11.7. Discussion

Chapter 12 - Multifaceted Nature of Data

12.1. More Noise-Free Data Usually Improves the Accuracy of AI Algorithms

12.2. Role of Bias in Data

12.3. Application Domains Where Bias in Data Is Particularly Harmful

12.4. Prevalent Methods for Mitigating Bias in Data

12.5. Data Ownership, Purpose, Consent, Privacy, Confidentiality, Auditability, and Governance

12.6. Creating and Using Synthetic Data

12.7. Discussion

Chapter 13 - Explainable, Interpretable, Causal, Fair, and Ethical AI?

13.1. The Need for Explainable AI

13.2. Important Techniques for Interpreting AI Models

13.3. Causality in AI

13.4. Fairness in AI

13.5. Ethical AI

13.6. Many Use Cases Do Not Require AI to be Explainable, Interpretable, Causal, Fair, or Ethical

13.7. Discussion

Chapter 14 - Maintaining and Improving the Accuracy of AI Systems

14.1. Managing AI systems is Significantly Different than Managing Traditional Software

14.2. Measuring Accuracy

14.3. Examples Where the Accuracy of AI Systems May Deteriorate

14.4. The Process for Maintaining Machine Learning Software

14.5. Maintaining Data Engineering Pipelines

14.6. Updating Machine Model Pipelines

14.7. Maintaining Other Systems Related to the Underlying Infrastructure

14.8. Maintaining Complete AI Systems & Continuous Governance

14.9. Recent Advancements to Mitigate Limitations and Improve AI Systems

14.10. Discussion

Chapter 15 - Future of Computing

15.1. Expected Demise of Moore’s Law

15.2. Three Characteristics of Quantum Computing

15.3. Key Algorithms and Advancements Related to Quantum Computing

15.4. Optical and Graphene Based Computing, DNA Storage, and Neuromorphic Computing

15.5. Proliferation of Special Purpose Hardware for Training AI Algorithms

15.6. Emergence of Special Purpose Hardware for AI Predictions and Inferences

15.7. Discussion

Chapter 16 - Jobs Likely to Be Lost and Gained During the Current Industrial Revolution

16.1. Jobs Likely to be Created and Lost Due to Ageing and Slowly Growing Population

16.2. Discussion of an Influential Article About Jobs Likely to Be Lost Due to Automation

16.3. Other Influential Articles Regarding Jobs Likely to Be Lost and Gained Due to Automation

16.4. An Analysis of Jobs Likely to Be Lost and Gained Due to Fourth Industrial Revolution by 2050

16.5. Jobs Likely to Be Created or Lost Due to Climate Change

16.6. Other Potential Job Gains in the Fourth Industrial Revolution

16.7. Discussion

Chapter 17 - Data, AI, and The Future of The Fourth Industrial Revolution (2023-2050)

17.1. The Imitation Game, Artificial General Intelligence, and Technological Singularity

17.2. The Future of AI May Rely on Neuroscience and Other Scientific Disciplines

17.3. First Three Characteristics of Industrial Revolutions Being Exhibited in the Fourth

17.4. Innovations Will Take Time to Seep in Society but Euphoria Will Cause Boom and Bust Cycles

17.5. Potential Roles of Government, Academia, and Private Investors

17.6. Discussion

About Dr. Alok Aggarwal

Appendix

A.1. Key Machine Learning Techniques That Are Being Used Today

A.2. Deep Learning Networks and Feed Forward and Backward Propagation Algorithms

A.3. Few Important Model Agnostic Techniques for Interpreting AI Models

A.4. Additional Details Regarding the Quantum Computing Model

Bibliography and Notes

Index

About the Cover

The movie, 2001: A Space Odyssey, starts with The Dawn of Man in which the black monolith appeared a few million years ago. The monolith appears mysterious and magical, and when one of the apes touches it, he seems to become innovative. He picks up a bone and begins to use it as a tool. First, he uses the bone as a tool to kill an animal. Then he uses the same bone as a weapon to kill a rival ape. In his exhilaration, he throws the bone in the air, and the movie begins to show the late 1900s with a rocket (also simultaneously a tool and a weapon) coming down instead of the bone.

Similarly, today AI is mystical and magical to most people. One of the aims of this book is to demystify AI and expose its achievements and limitations. And just like The Dawn of Man, it is now The Dawn of AI. The other aim is to point out how AI can be used as a tool or as a weapon (just like the bone or the rocket).

Introduction

Any sufficiently advanced technology is indistinguishable from magic.––––––––— Arthur C. Clarke

Even a month after seeing the movie, 2001: A Space Odyssey, some of its dialogs kept haunting me. During this movie, one person asks another, In talking to the computer (HAL 9000), one gets the sense that he is capable of emotional responses. For example, when I asked him about his abilities, I sensed a certain pride in his answer about his accuracy and perfection. Do you believe that HAL has genuine emotions? The other responded, Well, he acts like he has genuine emotions. Um, of course, he's programmed that way to make it easier for us to talk to him. But as to whether he has real feelings is something I don't think anyone can truthfully answer.

As a teenager in 1976 who had just finished one year of college, I kept wondering whether HAL 9000 was merely a fantasy or materially possible any time soon. Little did I know that many pioneers of Artificial Intelligence (AI) had been passionately working to build such a machine over the past two decades and had predicted that one would exist by the year 2000. For example, in 1961, Marvin Minsky wrote, Within our lifetime machines may surpass us in general intelligence, and in 1967 he reiterated, Within a generation, I am convinced, few compartments of intellect will remain outside the machine’s realm – the problem of creating ‘Artificial Intelligence’ will be substantially solved.

Demystifying AI – Although the field of AI was created over seven decades ago, it remains fantastical to many. In recent years, scores of commentators have epitomized AI as a mystic supernatural force – whether it be a champion for achieving utopia (AI might even save the world – Oren Etzioni, ex-CEO of Allen Institute for AI) or dreading it as a harbinger of doom (with artificial intelligence, we’re summoning the demon – Elon Musk, CEO of Twitter and Tesla). No wonder such hype has bolstered the view that the field of AI is more magic than science. Hence the first benefit of this book is to explain the science and engineering behind AI, which involves addressing the following eight questions:

What is AI, and what is its genesis? Chapters 2 and 3 discuss the evolution of AI.

What has AI achieved? Chapter 4 enumerates the essential achievements of AI.

What are the shortcomings of AI? Chapter 11 discusses key limitations of AI systems.

What enables accurate AI systems? Throughout this book, I show examples to prove how accurate AI can be and important characteristics that can lower accuracy.

If data is the enabler, then what are its key characteristics for enabling AI systems? Chapter 12 discusses limitations of data and its multifaceted nature.

What are good AI systems and what are their limitations? Chapters 11 and 13 describe limitations of contemporary AI systems and the progress as well as hurdles in building good AI systems.

How should we maintain and improve AI systems? Chapter 14 explores the challenges in managing and improving accurate AI systems.

How can we hope to improve AI systems in the coming decades? Chapter 15 discusses the limitations of classical computing and mitigating those limitations by using other technologies.

The Fourth Industrial Revolution – Just like the previous three industrial revolutions, the fourth revolution (which started in 2011) is expanding at a ferocious pace. For the last decade, the headlines have hyped prominent inventions that often went bust in the end. For example, in 2011, IBM Watson beat humans in the game of Jeopardy! and went through a boom-bust cycle during the nine years following that victory. Similarly, in 2015, Waymo demonstrated key inventions related to driverless cars. Between 2015-2020, pundits and investment bankers alike touted driverless cars, which led to a partial bust in 2021-2022. And during the last six months, one group of Deep Learning Networks in Artificial Intelligence called Generative Pretrained Transformers (GPTs) has captured human imagination worldwide. And while GPTs have been improving at an exponential rate, we still do not understand their capabilities or limitations.

Since the commingling of inventions of this revolution is immensely improving them, they are capturing headlines in the news and social media. This has resulted in a technological landscape that is mind-boggling with unclear implications. Given this backdrop, the second goal of this book is to discuss the vital characteristics of the fourth and current industrial revolution as well as its key inventions and their applications to society. Indeed, like the previous revolutions, this one will upend the status quo. For example, with gene editing and other healthcare inventions, the practice of medicine is likely to be transformed radically. Similarly, the inventions related to the newly created data infrastructure, AI, and climate change may end up simultaneously destroying and creating several hundred million jobs (discussed in Chapter 16).

In the first three industrial revolutions, steam engines, electric motors, and central processing units (CPUs) became diversified and ubiquitous. In fact, motors and CPUs are so widely used today (e.g., in washing machines, fridges, microwave ovens, phones, televisions, and computers) that they have almost become invisible. In the current revolution, by 2050 AI systems are expected to diversify analogously with innumerable uses in daily life. Similarly, in the first three revolutions, new infrastructures related to water and steam, electricity, and electronic communication were created. Correspondingly, the current revolution will lead to the creation of a new infrastructure related to ingesting, cleansing, harmonizing, and utilizing disparate datasets. In fact, these two iconic inventions (i.e., AI and novel infrastructure regarding data) will also improve the following inventions of the current revolution:

Internet of Things (IoT), which is discussed in Chapter 5.

Inventions related to predicting, mitigating, and adapting to rapid climate change, which are explored in Chapter 6.

Blockchains, which are explained in Chapter 7.

Metaverse and its potential applications, which are elaborated in Chapter 8.

Robotics, driverless vehicles, and three-dimensional printing, which are described in Chapter 9.

Inventions related to gene editing, protein folding, and healthcare, which are stated in Chapter 10.

Undoubtedly, each industrial revolution has had an enormous impact on society by affecting the workforce, the role of governments, or driving the trajectory of science (these are discussed in Chapters 16 and 17). Furthermore, to bolster the arguments provided in Chapter 17 (that AI systems will be used in numerous applications), more than 1,000 applications are listed on www.scryai.com. Lastly, although comprehensive with additional technical details in the Appendix, this book contains very little math and no software code.

Primary Audience – Overall, this book aims to provide crucial information to the following:

Students, especially graduate students who are in science, technology, engineering, mathematics, analytics, business administration, financial engineering, and related disciplines. Each of the previous revolutions lasted for four or more decades, and the current one is likely to be no different. On the other hand, during the next ten to twenty years, many current students will become entrepreneurs and decision-makers in diverse organizations. Hence, they would be ideally suited to exploit these inventions, many of which would have started seeping into society.

Product managers and program leaders who may not need to understand the minute details of AI systems but should have sufficient knowledge to discuss with clients and internal technology teams.

Business leaders who wish to understand AI at a broad level and use it to improve their organization’s processes.

Consultants and investment managers who advise their clients and need a general understanding of AI. These people can use AI to improve their business processes or for starting or acquiring other businesses.

Keeping this discussion in mind, the next chapter begins with a discussion regarding eight significant characteristics of three previous industrial revolutions and how these characteristics are already exhibiting themselves in the current one. Indeed, these characteristics are useful in understanding the pace and scope of the current revolution and in imagining the possibilities that may unfold during the next three decades.

Chapter 1

Vital Characteristics of The Ongoing Fourth Industrial Revolution

In the late 1890s, Thomas Edison was developing a nickel-iron battery when his friend, Walter Mallory, visited his laboratory and asked, Isn't it a shame that with the tremendous amount of work you have done, you haven't been able to get any results? To this Edison quipped, Results! Why, man, I have gotten lots of results! I know several thousand things that won't work [101]. Given that Benjamin Franklin and others discovered electricity in the 1750s and Michael Faraday showed how to produce electricity in the 1830s, it still took innovators several decades to invent new gadgets for humans to use electricity. Indeed, such innovations and their commercialization would not have been possible without inventors’ relentless pursuit to innovate. And as discussed throughout this book, this feature constitutes one of the hallmarks of all scientific and industrial revolutions.

During the last three hundred years, the world has witnessed three industrial revolutions. As will be discussed in this chapter, all these revolutions had eight characteristics in common. In December 2015, Klaus Schwab mentioned that we are amidst the fourth revolution [102]. Throughout, this book provides ample evidence that since 2011, these eight characteristics have been exhibited by society and the scientific community, thereby implying that the fourth revolution began in 2011 and is continuing vigorously. Unsurprisingly, these characteristics will manifest even more during the next few decades.

This chapter is organized as follows. Section 1.1 briefly discusses key inventions of the current industrial revolution which will be discussed later in detail. Since all industrial revolutions are based on scientific discoveries, it is important to understand the structure of scientific revolutions which is discussed in Section 1.2. Sections 1.3 through 1.8 discuss eight shared characteristics of previous industrial revolutions, and how these characteristics are exhibiting themselves in the current revolution. The first three characteristics are discussed in Section 1.3. Two of these characteristics were vital for each industrial revolution because one led to the creation of a new infrastructure, whereas people used another pervasively. Section 1.4 contends that it usually takes a substantial amount of time for scientific innovations to percolate through human society. However, since revolutions usually create euphoria and hype, Section 1.5 discusses the boom-and-bust cycles that often occur during the process. Section 1.6 argues that once these key inventions seep into society their effects are significantly more than anticipated by their inventors. Section 1.7 examines how these revolutions upended the status quo and created new jobs while destroying older ones, thereby impacting society immensely. The role played by various governments in these revolutions is the eighth characteristic and is discussed in Section 1.8. Finally, Section 1.9 concludes by depicting the interplay between scientific and industrial revolutions and their implications for the current industrial revolution.

1.1. Key Inventions of The Current Industrial Revolution

Each previous industrial revolution is marked by the rise of several new inventions, and the existing revolution is no different. Given here are the key inventions of the current and fourth industrial revolution:

Generation, Distribution, Cleansing, Harmonization, and Use of Data – Although data is multifaceted and markedly different than electricity (to be discussed further in chapter 12), its infrastructure is like that of electricity, which is produced at generating stations, transmitted via wires and cables, modified at sub-stations (e.g., by reducing voltage), and then used by organizations and individuals. Similarly, datasets are produced by a plethora of sources, transmitted via electronic communication, stored in personal devices or those that are on the Internet or Intranets, cleansed, transformed (e.g., harmonized with other datasets), and then used for improving workflows, processes, and other inventions.

Artificially Intelligent (AI) Systems Becoming Pervasive – AI systems try to mimic non-trivial human tasks with high accuracy. Although AI systems were invented in the 1950s, their rampant commercialization only became possible in 2011. And today, AI systems are analogous to electric motors which come in numerous forms and sizes.

Internet of Things (IoT) – IoT includes sensors and devices that collect and transmit data typically via the Internet or Intranets. For example, a video camera and an internal fire sprinkler may cover a portion of an office to collect data and then send that data to a computer that can determine if there is a fire in that office.

Inventions Related to Predicting, Mitigating, and Adapting to Climate Change – Undeniably, the rapid rate of climate change will be catastrophic for human society, especially if it is left unchecked. Hence, several inventions are coalescing around predicting, mitigating, and adapting to rapid climate change.

Blockchains – Instead of being centralized systems like banks, Blockchains are comprised of decentralized systems for conducting financial and non-financial transactions where the entire community is in charge (rather than a few entities), and where all transactions are immutable, auditable, and transparent. The digital currency related to financial blockchains is called cryptocurrency.

Metaverse – Virtual Reality (VR) is a simulation of a three-dimensional world that people experience after wearing special headsets or similar equipment. The computer gaming and entertainment industries are already using VR. In its simplest form, Metaverse extends this concept to allow buying, selling, and renting of virtual real estate.

Robotics, Driverless Vehicles, and Three-Dimensional (3D) Printing – Substantial improvements with respect to Robotics, autonomous vehicles, and 3D printing have already occurred, and many more will occur during the next ten to fifteen years. For example, 3D printing is already being used to build new Robots.

Key Inventions Related to Healthcare – Researchers are discovering new and exciting ways to edit genes, fold proteins, and discover new drugs, a lot of which are being powered through AI.

Quantum, Graphene, Photonics, and Other Computing Methods – Since we are reaching physical limits regarding the underlying hardware, researchers are pursuing other avenues for complementing classical computing. Most promising among these include Quantum, Graphene, and Photonics computing.

Eventually, all industrial revolutions are built on scientific inventions and discoveries. Accordingly, these industrial revolutions can sometimes slow down if the corresponding advances in science and engineering have stalled. Hence, scientific revolutions are required to get out of this quagmire, and their structure is briefly discussed next.

1.2. The Structure of Scientific Revolutions

A revolution refers to a fundamental change in the way of thinking about or visualizing something, a change of paradigm, a changeover in the use or preference, especially in technology, or a movement designed to effect fundamental changes in the socioeconomic situation.

In 1962, Thomas Kuhn published a book titled, The Structure of Scientific Revolutions [103]. In his book, Kuhn provided a thesis about how scientific revolutions occur, which is briefly discussed here, and which will be discussed within the context of AI in Chapters 2 and 17.

Science Works with Paradigms: In Kuhn’s viewpoint, normal science works within paradigms. These are philosophical and theoretical frameworks of a scientific discipline within which theories, laws, and generalizations are formulated and experiments performed. Roughly speaking, paradigms are theoretical belief systems followed by scientists at any given point of time.

The Process of Normal Science and Occurrence of Anomalies: Once a group of scientists begins to follow a paradigm, Kuhn believed that normal science aims at expanding the current paradigm. However, at some point in time, this expansion ends because one or more anomalies are observed. A crisis occurs when several anomalies become inexplicable and lead to an explicit discontent ... and the debate over fundamentals. In other words, the status quo with the current paradigm (i.e., with the current theoretical belief system) does not help in resolving these anomalies, and the quest begins to find a new one.

New Paradigms Are Adopted: Eventually, scientists provide new ideas that lead to the creation of new paradigms. Once the new paradigms begin to resolve current anomalies and paradoxes, they foment a revolution by creating a fundamental change in the way scientists think. Soon, many begin to follow the new paradigm (i.e., the new theoretical belief system). After which, these steps are repeated.

For example, in Physics, the Copernican Revolution helped Copernicus, Galileo, Newton, and others in creating a new paradigm. Roughly speaking, this paradigm implied that an entity can be either matter or a wave but not both. This paradigm worked well and was even expanded for two centuries but eventually ended in anomalies and paradoxes that confounded Physicists in the 1880s. In the early 1900s, Albert Einstein provided special and general theories of relativity, and Max Planck introduced quanta which led to the field of Quantum Physics. These new paradigms implied that in various experiments, an entity can demonstrate properties related to matter at one point of time and that of a wave at another. Indeed, these new paradigms were so radical that they not only toppled the Copernican paradigm, but they even shook Einstein up, who stated, God does not play dice with the universe. A sentiment to which Neils Bohr retorted, Einstein, stop telling God what to do with his dice [104].

Regarding the fourth industrial revolution, such a scientific revolution occurred in the 1950s that led to the genesis of Artificial intelligence. This genesis will be discussed in detail in Chapter 2. Furthermore, Chapter 17 contends that although research in AI has made significant strides and is now being commercialized enormously, the most general form of AI may not be achieved without another scientific revolution.

1.3. The Structure of Industrial Revolutions and Three Key Characteristics

During the last three centuries, three industrial revolutions occurred between 1760-1840, 1870-1914, and 1950-2010 respectively. A good way to understand the fourth revolution and its characteristics is to give examples of these characteristics from previous revolutions. Hence, this section discusses the following three traits and their distinct presence in each of these revolutions.

One Key Invention Led to Creating a New Infrastructure – In each industrial revolution, within a few decades, one invention led to the creation of a new infrastructure which became an integral part of the infrastructure and human society.

Cambrian Explosion of Another Key Invention – Almost simultaneously, in each industrial revolution, another invention had its own Cambrian explosion and became ubiquitous. The Cambrian explosion of living species started approximately 540 million years ago and lasted for roughly twenty million years. Before this period, most organisms were individual cells or were small and multicellular in nature. During the Cambrian period, diversified and complex multicellular organisms started appearing in large numbers, and the variety of life became substantially more complex and began to look like it is today. Like the Cambrian explosion, each industrial revolution had its own equivalent with one invention becoming abundantly diversified and appearing in numerous forms, structures, and sizes.

Each Revolution Was Marked by Several Key Inventions – Many inventions constituted each revolution, and many others were created almost simultaneously and in conjunction with each other. Moreover, these innovations were created several decades after the corresponding scientific revolutions in Physics, Chemistry, and other sciences because innovating them was laborious and time-consuming.

The First Revolution (1760-1840) took place primarily in Britain. The two iconic inventions of this era included (a) the generation, distribution, and use of steam and (b) the introduction and diversification of steam engines. By the end of this revolution, the invention related to efficient generation and distribution of steam had become a part of the enabling infrastructure, and most inventions in Britain and elsewhere used steam power.

In 1776, James Watt markedly improved the steam engine that was developed earlier by Thomas Newcomen [105]. The new steam engines were so efficient that they began to be used in conjunction with other inventions, thereby transforming some of the largest industries (e.g., textile manufacturing, machine tools, cement, chemicals, flour, paper, distilleries, waterworks, and canals). Due to the inventions by Trevithick, Stephenson, and others in the early 1800s, steam-powered locomotives started transporting passengers and freight, and steam-powered boats and ships began carrying goods across canals, rivers, and the seas. By the end, steam engines had their equivalent of a Cambrian explosion and became so pervasive that including them in other inventions was a cakewalk.

Additionally, key inventions of this revolution included the spinning jenny, water frame, and the spinning mule. James Hargreaves, Richard Arkwright, Samuel Crompton, and several others kept improving these inventions for decades until they made textile manufacturing more efficient and less laborious [106]. Notably, all these inventions relied on the fundamentals that were created during scientific revolutions, specifically the Copernican revolution several decades earlier.

A drawing of a machine Description automatically generated

Figure 1.1: James Watt's rotative steam engine, 1788 [Science Museum, London]

The Second Revolution (1870-1914) was led by the United States and Europe. The two most widespread inventions of this era included (a) the generation, distribution, and use of electricity and (b) the introduction as well as the diversification of electric motors.

The backbone of this revolution was the generation, distribution, and use of electricity. By the mid-1920s, this infrastructure had become a vital part of society’s infrastructure and people had started using electricity in numerous ways.

Simultaneously, electric motors had a Cambrian explosion and began to be built in diverse forms, shapes, and sizes. Today, motors are everywhere and usually invisible. In fact, their worldwide market is expected to be 220 billion US Dollars by 2030 and they are so deeply embedded that including them in a new gadget is a no-brainer [107]. For example, motors are being used in heating, ventilation and air conditioning, industrial automation, agriculture, compressors, blowers, fans, refrigeration, crushers, lathes, drills, power tools, rolling mills, paper mills, conveyors, washing machines, drying machines, elevators, escalators, computer disk drives, printers, and photocopiers, positioning and heavy equipment, hoists, winches, and Robots.

A black and white picture of a machine Description automatically generated

Figure 1.2: Electric generator in the 1880s

Other Key inventions included electric trams and railways, the telephone, the telegraph, the incandescent lamp, the internal combustion engine (which replaced the steam engine and fueled the automotive industry), the QWERTY typewriter, the automobile, and the mass production of consumer goods including vehicles [108]. To perfect these inventions, Michael Faraday, Werner Siemens, Alexander Bell, Samuel Morse, Nikolai Tesla, Thomas Edison, and many others kept improving them for decades. Also, just like the first industrial revolution, all inventions during this revolution relied on the first revolution in Physics and the fundamentals discovered by Benjamin Franklin and others several decades earlier.

The Third Industrial Revolution (1950-2010) started after the Second World War in the United States and quickly spread to Europe and other parts of the world. The two most widespread inventions of this era included the abundant use of electronic (satellite, wireless, and wireline) communication and the diversification of central processing units or CPUs.

Just as the generation and distribution of electricity became the enabling infrastructure in the second revolution, electronic communication became part of the society’s infrastructure in the third one and began to power many other inventions.

Similarly, just like motors in the second revolution, central processing units (CPUs) had a Cambrian explosion. They became pervasive and started coming in various shapes, forms, and sizes. For example, firms started embedding them in other electronic and mechanical products including main-frame computers, personal computers, mobile phones, video cameras, vehicles, and numerous electronic devices and sensors.

A close-up of a black cell phone Description automatically generated

Figure 1.3: Blackberry 850 in the 1990s

Other key inventions of this revolution included the first electronic general-purpose computer – The Electronic Numerical Integrator And Computer (ENIAC) – in 1946 [109]. Shockley, Bardeen, and Brattain developed the first silicon-based transistor, which was substantially improved by Atalla and Kahng to become the fundamental building blocks of digital electronics and central processing units or CPUs [110]. Also, Marconi’s original invention of wireless communication in the late nineteenth century was substantially improved to create digital wireless networks in the 1990s. In 1973, Martin Cooper and his colleagues built the first handheld mobile phone [111]. Chapin, Fuller, and Pearson developed the first photovoltaic cells in 1954, which was almost five decades after Einstein had published his research on the photoelectric effect. The Defense Advanced Research Projects Agency (DARPA) created ARPANET which was converted to the Internet and paved the way for Tim Berners-Lee to propose the World Wide Web in 1989 and develop an initial version by 1990 [112].

The Fourth Industrial Revolution Started in 2011, and like other revolutions, it may continue for 40 years or more. Furthermore, the three characteristics mentioned here are already manifesting themselves in this revolution. More specifically:

The production, communication, cleansing, transformation, and consumption of data are leading to the creation of a new infrastructure (to be discussed further in later chapters).

AI systems are already being used pervasively and are likely to have their Cambrian explosion during the next eight to ten years.

As mentioned in Section 1.1, this industrial revolution also comprises many key inventions that are already benefiting from each other.

1.4. Technology Takes Substantial Time Before Becoming Pervasive

The fourth characteristic is as crucial as the three mentioned here because we often forget that most game-changing inventions took several decades to become widespread in society. For example, during the first industrial revolution, Trevithick built the first steam locomotive in 1804. Stephenson improved it in 1830 and used it for the first public railway system between Liverpool and Manchester [113]. However, railways became prevalent in Britain and the United States only four to six decades later.

Similarly, during the second industrial revolution, although the Edison Electric Illuminating Company started providing electricity to parts of New York in 1882, it was only around 1925 that half the homes in the United States finally had electricity [114].

Listed here are a few reasons why human society often takes substantial time to fully integrate even the most vital inventions.

Large Capital Investment Is Required: Many inventions need substantial infrastructure improvements and capital infusion. For example, railways required several hundred thousand miles of railroads to be built during the second revolution. Similarly, several million miles of broadband fibers needed to be installed during the third revolution.

Need for Obtaining a Return on Past Investment: Often, existing companies and consumers have already invested in older technology and want to extract their return on investment before investing in a newer one. For example, many people who bought gasoline (petrol) powered cars recently may not buy new electric vehicles immediately.

No Urgent Need to Fix the Current Process: Organizations usually feel that if it ain’t broke, don’t fix it, which is the main reason why almost 3% of the global economy still runs on software programs built using COBOL language that is dead and almost impossible to upgrade.

Massive Resistance to Change the Current Business Model: Often, significant inventions require business models to be changed, which companies loathe to do because they believe, What got me here will also get me there. Hence, key inventions often require new firms – generally startups – with new business models to be created and to grow, all of which take time.

Risk Aversion: Most companies and consumers are usually risk averse and concerned about being blamed if an invention fails to perform adequately. Hence, they are not prone to adopt a critical innovation unless they have observed it work for others.

Need to Retrain Workforce: Many inventions require the workforce to be retrained or upskilled. For example, the ability to read, write, and understand user manuals became more important as inventions in the second and third revolutions became common.

New Government Regulations May Be Required: Frequently, to accommodate a critical invention, government regulations need to change, which is time-consuming.

Consumers Take Time to Adapt: Innovations require consumers to adapt appropriately. This adaptation is time-consuming, especially for older people.

Robert Gordon emphasizes the fourth trait eloquently [115] (see Figure 1.4). Even though, from 1900-1915, human society was amidst the second revolution and various game-changing inventions were pushing to go mainstream, productivity languished during these fifteen years. In contrast, productivity growth was much higher between 1930-1970 when most inventions of the second revolution had seeped into society.

Figure 1.4: Annual productivity increase in the United States [115]

Implications for the Current Industrial Revolution – All the reasons mentioned here are likely to impede the pace of the current industrial revolution. For example, our analysis shows that creating a gigantic infrastructure regarding production, communication, cleansing, harmonization, and usage of even 5% of all existing Internet data is likely to take more than 5 trillion US Dollars. In fact, creating such an infrastructure for a large bank like Citibank is likely to take more than 5 billion US Dollars. Indeed, firms of all sizes will loathe to invest so much money unless they see a return on investment in a relatively short period. Fortunately, just like local electric generators can often fulfill the demands of even manufacturing companies, various organizations may not always require a gigantic infrastructure for data. In fact, for the next five to ten years, many firms are likely to reap benefits by creating a limited data infrastructure to improve their workflows and processes, thereby obtaining their return on investment.

1.5. Hype May Be Irrational Exuberance But Can Be Advantageous

Even the most vital inventions take time to seep into society. However, inventors, investors, and others usually become overly euphoric about these inventions. Such euphoria is the fifth shared trait among industrial revolutions, which leads to boom-bust cycles.

On one hand, exhilaration induces the innovators to achieve astonishing feats. On the other, it leads them as well as investors into believing that their stupendous inventions are so remarkable that they will seep into society almost instantaneously. Hence, they fool themselves and others into believing that this time, it is different, and they forget the lessons mentioned in Section 1.4 that most game-changing inventions take several decades before they are commonly used by the masses. In fact, such misconception is so rampant that even most think tanks, strategy companies, and people running businesses often underestimate – by a factor of two or more – as