1. The Technological Basis of Breakthrough Disruption – From Incremental to Exponential

ONE

The Technological Basis of Breakthrough Disruption

CHAPTER SUMMARY: This chapter covers what happens as information becomes widely and cheaply available. Using how Uber benefited from Google Maps and soaring smartphone usage, we discuss how the next wave of disruption is coming from artificial intelligence (A.I.) making good predictions cheap or free. The chapter details how the rapidly falling costs of technologies such as computer chips, sensors, and network capacity, are increasingly enabling broader possibilities, such as the ubiquity of the smartphone.

To obtain a license to pilot a black cab in London, applicants must pass an incredibly difficult exam called the Knowledge. The exam requires them to show that they have memorized the Byzantine street layout of the British capital and can calculate the most efficient path from point A to point B at any given time of day. Navigating London’s chaotic and unpredictable arterial streets and congested roundabouts is challenging, and drivers able to demonstrate such ability were paid relatively well, considering that this was not an advanced professional degree. This made the job of driving a black cab highly desirable.

Then Uber entered the market. The prominence of black cabs entered a steep decline as Uber grabbed market share through its cheaper fares and its ubiquity. But a quick look under the hood reveals something more: Uber didn’t disrupt black cabs all by itself. Google Maps played an equally central role in its demise. Google Maps effectively made the Knowledge available—for free—to anyone with a smartphone, and Uber initially used Google Maps for turn-by-turn navigation for its drivers. It is likely that, had it arrived without a means of minimizing the importance of the Knowledge, Uber would not have succeeded.1

By making available for free (or nearly free) cartographic information that had been expensive and hard to obtain, Google affected a wide variety of companies that depended on the value of geographical information, from navigation device makers such as Garmin and TomTom to the sellers of geographic data such as Telenav. As an indication of the disruption that Google Maps and the emergence of free turn-by-turn apps on smartphones caused: Garmin’s market capitalization, which in September 2007 was more than $16 billion, went into freefall when Google Maps became available, plummeting to the $2 billion range. It took Garmin a long 12 years to regain a $16 billion market capitalization,2 which it did by transforming its entire business model and locating alternative revenue sources beyond its formerly dominant turn-by-turn navigation systems.3

This pattern is common to breakthrough disruptions. Unlike the days in which the disruption came from a cheaper product’s crashing the market at the lower end, the new form of industry disruptions makes key business activities nearly free or incredibly cost effective, enabling upstarts to very quickly enter and capture market share or to build entirely new businesses on the new economics these disruptions enable.

In many fields, accurate predictions were formerly very expensive, if not impossible. Predictions are becoming far more affordable and often free—through the analysis of mountains of data that we already have, through the use of technologies such as A.I. As economists Ajay Agrawal, Joshua Gans, and Avi Goldfarb explain in their book Prediction Machines: The Simple Economics of Artificial Intelligence, freely available predictions will fundamentally change how we conduct our lives and how business behaves:

Having better prediction raises the value of judgment. After all, it doesn’t help to know the likelihood of rain if you don’t know how much you like staying dry or how much you hate carrying an umbrella. Prediction machines don’t provide judgment. Only humans do, because only humans can express the relative rewards from taking different actions. As A.I. takes over prediction, humans will do less of the combined prediction–judgment routine of decision making and focus more on the judgment role alone.4

Although A.I. is (contrary to popular belief) nowhere near replacing human intelligence, free predictions will enable the humans who embrace its capabilities to make faster and more effective decisions.

The case of Uber, black cabs, and the Knowledge made a clear analogy a decade before A.I. began making predictions free. Geographic knowledge at any significant scale before the rise of computing, satellites, cheaper sensors, ubiquitous connectivity, and high-speed wireless networks was dear. Further back in time, in the days of Christopher Columbus and Amerigo Vespucci, the maps that held geographic knowledge determined the wealth of nations. Today, knowledge is becoming far less expensive, and more knowledge is accumulating at a dizzying pace. This accumulation would be crippling if not for the rise of A.I. With A.I., we can make sense of much of this noise.

Thus, A.I. and its impact on predictions—and subsequently on business—are the tip of the spear of exponential disruption.

Sensors, Chips, Software: How Exponential Technologies Combine to Become Breakthrough Innovations

The common perception that the technology world is moving ever faster is not wrong: adoption of newer generations of technology is occurring more quickly today than fifty or even twenty years ago. Earlier technologies with the potential to transform our world, such as the steam engine and electricity, required as long as a century to attain widespread use. Radios and televisions penetrated more quickly than electricity did, but widespread adoption of each took several decades. From the emergence of the computer to nearly 90 percent penetration took about two decades. The smartphone’s far more global penetration took a single decade. The adoption of newer technologies, such as voice assistants, is even faster, with significant uptake occurring in roughly five years. A.I. too, although it had been bubbling in the background for decades, went from nascent to nearly omnipresent in roughly five years.

So yes, everything is moving faster, and the pace of change is accelerating. A.I. will help us understand what is going on around us and to make predictions, making a scarce good or service free. The cost of clean energy will fall to a point at which it seems free. Everyone on Earth who wants an Internet-connected smartphone or V.R. headset will have one, giving us a truly global span for information sharing. The platforms supporting such technology are also getting faster. Emerging 5G wireless networks, modern WiFi, and giant fleets of cheap communication satellites launched into low-Earth orbit will provide fast broadband everywhere for a fraction of the current costs of electronic communication. The scramble for A.I. is driving a new wave of computer-chip startups seeking to address problems of machine learning, and the resulting chip designs hold tremendous promise for performing computing tasks far more efficiently and elegantly than occurs now. Impediments to innovation and rapid technological acceleration are falling away as speeds rise, costs fall, and adoption quickens.

The human genome is a case in point. Sequencing the first human genome cost roughly $2.7 billion: a tremendous effort involving large teams of scientists and labs full of equipment. This occurred in the year 2000, as a result of a thirteen-year government-led effort.5 Today, some labs can fully sequence a human genome for less than $1,000; the cost will fall below $25 in less than a decade, as a result of improved technologies for DNA analysis: a highly automated process running on laboratory tools that are essentially high-speed computers using cheap sensors to prise open and analyze the formerly mysterious double helix that is the recipe book of life.

What Is Driving Exponential Innovation

In their book The Driver in the Driverless Car, Vivek Wadhwa and Alex Salkever detail a range of technologies that are advancing on an exponential curve and the possibilities they have enabled. These diverse technologies are also converging. This convergence, or combination, creates opportunities for entrepreneurs to disrupt entire industries.

You have seen the advances in our computers, how they keep getting faster and smaller. The Cray supercomputers of the 1970s were considered strategic government assets. They could not be exported; they were for scientific research and defense; and they cost tens of millions of dollars. They needed to be housed in huge buildings and required water cooling. The smartphones many of us carry in our pockets are many times more powerful than the Crays were.

This progression follows an industrywide development cycle known as Moore’s law. For more than half a century, the speed, efficiency, cost-effectiveness, and power of computing devices has doubled roughly every 18 months. Faster computers are now used to design even faster computers; and computers—and the information technology (I.T.) they enable—are absorbing other fields. The result is exponential advances in sensors, A.I., robotics, medicine, 3D printing, and more. To paraphrase futurist and inventor Ray Kurzweil, as any technology becomes an information technology, it starts advancing exponentially.

The advances in sensors such as the camera on a smartphone are illustrative. Kodak introduced the first “computerized” camera in 1976. It weighed 4 pounds, cost $10,000, and had a resolution of a whopping 0.01 megapixels. Today, the cameras in some mobile phones have 108-megapixel photosensors, and the cameras are merely an extra. You may remember, too, the really expensive high-definition cameras that film studios started using at the turn of the century. Apple’s iPhones from the iPhone 9 Plus to the iPhone 11 shoot video at four times the resolution (3,840 × 2,160 pixels) of those early professional cameras. Similar advances have occurred in other types of sensors: in accelerometers; gyroscopes; and sensors of temperature, gas, humidity; as well as in microfluidics6 and in the performance of chemical and biological tests on small, inexpensive chips.

Sensors’ improvements in accuracy and reductions in price are also facilitating a revolution in manufacturing, in which their rapid spread (the “Industrial Internet of Things”) has enabled what many now term Industry 4.0, rapidly raising manufacturing processes’ efficiency by capturing data on every key variable of production: pressure, temperature, ambient humidity, percentage of scrap, precision of casts—you name it. This information gives factory processes digital souls that are transparent and tunable, almost like software code. Like the smartphone, Industry 4.0 became possible when multiple elements—sensors, connectivity, computing power—became cheap, powerful, and compact.

It’s when exponentially advancing technologies combine that the magic happens. Such convergence makes possible new applications and allows the creation of new industries at the cost of the older ones.

If we take the data we are collecting from sensors, the Internet, and the computerization of almost all knowledge work and apply A.I. to their analysis, we obtain the ability to predict traffic patterns, crime, sales, and trends.

Computers will also soon be performing medical diagnoses. The Apple Watch and Fitbit are medical devices that use advanced sensors to monitor our health. Thousands of such medical sensors are in development worldwide. They will monitor our activity levels and our sleep; our vital signs and body fluids: everything about us. A.I. systems in the form of smartphone apps will read these data 24/7. They will advise us when we are about to fall ill and will recommend better lifestyles, habits, and treatments. These technologies are becoming possible because of the combination of sensors, computing, medical libraries, and A.I. We now even have sensors, such as sugarBEAT from Nemaura, entering the market that will allow non-invasive glucose monitoring, changing the lives of hundreds of millions of diabetics in the world. In a decade or so, we will not need doctors to advise us on day-to-day health; their work will be for the complicated ailments. These incredible technologies will disrupt the entire medical industry.

The Limits of Technology

When we were young, many of us watched TV shows such as Star Trek and dreamed of replicators—which would produce all the icecream and dessert we could eat—or of having a robotic assistant like Rosie from The Jetsons clean up after us. But Rosie never came, and all we’ve had by way of replicators are 3D printers that print cheap plastic toys. Indeed, the most advanced robots in our homes today are Roombas—pathetic little automated vacuum cleaners. (Yes, we know that some people love their Roombas!)

Why No Rosie?

There is no Rosie because the computation power required for a robot to recognize voices and speak intelligently would have required a Cray computer, and the sensors—camera, motion detectors, gyroscopes, accelerometers—were too bulky and expensive.

Guess what a smartphone can do today, though: all of that, and much more.

Rosie has become possible, and it is conceivable that Amazon will deliver her to our homes by drone in the late 2020s. We’ll also see robots doing the jobs of humans in manufacturing plants, grocery stores, and pharmacies. And they will be driving cars and making deliveries. Robots will soon do all the routine things that humans do. Imagine the resulting possibilities and disruptions.

Replicators too are on their way. There are already demonstrations of “3D-printed” meat and desserts by startups in many countries. We will be 3D printing not only food but also cars, electronics, houses, and space stations.

Every industry in which technology can be applied or that generates data faces these advances. There may be no industry whose leading players won’t face economic extinction.

It is important to acknowledge that, although advancing technologies enable a lot of good, they also enable large-scale destruction, spying, and unimaginable horrors. Already creating social and ethical dilemmas, they are taking us into a future in which there won’t be much work in the professions of today and we’ll have to figure out what to do with ourselves. There will surely be social unrest as the rate of change accelerates and the gap between the haves and have-nots widens; there will also be efforts to halt the progress of certain technologies. If we are to be masters rather than victims of our tools, this ever-present choice of futures (the key message of The Driver in the Driverless Car) is one we must come to grips with.