The Age of Big Promises and Small Results For New Technologies
We live in an era of big promises but results that fall far short of expectations. Take the failed construction startup Katerra. Founded in 2015, the company claimed it would use the approaches of digital and mass production industries, including glued and laminated “mass timber” products and modular design, to “disrupt” the construction industry, long seen as a bastion of backwards, inefficient craft labor. Many people enthusiastically bought into the vision of Katerra. Over six years, it raised $3 billion, including $2 billion from Japanese telecom and investment giant SoftBank Group.
In 2021, the global construction industry had a market size of about $11.5 trillion. Any company that truly disrupted and claimed a large chunk of such a huge market would be one of the largest winners in industrial history. But Katerra was not to be that company. It failed for many reasons, including because it hubristically underestimated the complexity of construction. It went belly up in June 2021.
Journalism and popular culture today are full of similarly large promises around a slew of other emerging technologies, including artificial intelligence, self-driving vehicles, the sharing economy and delivery apps, AI for healthcare, fusion energy, nanotechnology, bio-electronics, virtual reality, augmented reality, commercial drones, domestic robots, blockchain, Internet of Things, asteroid mining, and smart homes. Yet, so far at least, the hype around these technologies far outweighs the creation of successful products and services and growing markets for them.
There is plenty of countervailing evidence that should lead us to question the dramatic claims made of these technologies and the health of an economy so deeply and highly invested in the “tech sector.” Yet the many setbacks and outright failures rarely lead industry leaders and boosters to reflect, instead of merely upping the promises. It’s now very well known amongst those who follow the economics of technology that the United States and other industrial nations have experienced pronounced low productivity growth since 2004, precisely the period during which we have heard some of the most exaggerated claims about some of the technologies listed above. Moreover, one of us (Funk) has published several articles demonstrating a gap between hype and reality: one, for example, showed that artificial intelligence technologies are unlikely to produce significant productivity changes soon and will more likely give birth to slow incremental improvements over decades. Another showed that annual and cumulative losses for today’s startups are far higher than those of previous decades, suggesting big problems in venture capital. If these new technologies are so great, why can’t they make money?
More recently, however, we, the authors, have found that the gap between promises and reality become even clearer when we compare much-hyped technologies of the last decade to new technologies of previous decades that did lead to significant change. To show just how ridiculous are recent claims, we gathered data on the revenues of new technologies from previous decades, with an emphasis on new ones that are purportedly behind the so-called digital transformation of companies, factories, homes, and roads. We find a fairly radical disconnect between newer technologies and what came before.
Most observers would consider 1950s computing and their electronic components, transistors, and integrated circuits, to represent the beginning of the digital transformation. Mainframe computers began to diffuse in the 1950s and 1960s, mini-computers and robots in the 1960s and 1970s, and personal computers in the 1970s and 1980s, with packaged software for them not far beyond. Behind these new computers were rapid improvements in microprocessors (from late 1960s), memory (from early 1970s), and graphic processors (from early 1980s). These changes developed enormous markets. For example, in 1989, 21 million PCs were sold at about $3000 each for a total market of $63 billion or $132 billion in 2020 dollars.
Bigger changes began to occur in the 1990s as networking equipment enabled these computers to be connected both within and between companies, largely based on rapid improvements in fiber optics. The Internet was born. Construction of the Internet accelerated, giving us e-commerce, enterprise software such as customer relationship management and manufacturing resource planning, and widespread use of mobile phones. These changes also quickly led to large markets. E-commerce, Internet hardware, and software, and mobile service revenues had reached $446, $315, $282, and $230 billion respectively by 2000 (1998 for mobile services), all in 2020 dollars to simplify comparisons to subsequent decades. Internet-connected personal computers also likely led to significant economic growth, with a period of high productivity gains between 1994 and 2004 that outpaced both the period from 1970 and 2004 and the period between 2004 and the present.
The 2000s were the beginning of rapid growth for smartphones, cloud computing, online advertising, social networking, and e-books. Cloud computing had global revenues of $127 billion by 2010 (also in 2020 dollars), and online advertising of $81 billion by 2010. Facebook had 550 million users by the end of 2010. (Some of this diffusion accelerated during the end of the 2000s and thus the quoted figures are for a year after 2010.) The iPhone was introduced in 2007, the app store in 2008 and Android phones also in 2008. The global revenues for smart phones had reached $293 billion by 2012 and web browsing, navigation services, and other apps were widely used.
We see something shifting in the 2010s, which were a decade of growing markets for existing technologies, but less so for new ones. Although revenues for e-commerce, cloud computing, smart phones, online advertising, and other technologies mentioned above continued to grow, only one single category of new technology had achieved $50 billion in sales by 2020, and that was video streaming which is more applicable to consumers than to corporate digital transformation efforts. (See the table below.) The next closest was Big Data/Algorithms with $46 billion, tablet computers with $40 billion (iPad introduced in 2010), and OLED Displays with $32 billion in revenues; the latter is also not usually considered part of corporate digital transformation efforts. Artificial Intelligence, virtual reality, augmented reality, commercial drones, smart homes, and blockchain have even smaller markets. Even the Internet of Things had only reached 1/5 the number of connected devices projected in 2012 for 2020 and most of those devices were smart phones. To add insult to injuries, many hyped startups in these sectors, such as Uber, Lyft, Palantir, Airbnb, Bloom Energy, Nutanix, and Snap have cumulative losses of more than $3 billion each, the amount Amazon had at its peak, they are still unprofitable, and their total cumulative losses are $58 billion and rising.
Again, contrast these market sizes with what came before: E-commerce, which was launched with the commercialized internet in the early 1990s, had reached $446 billion (in 2020 dollars) less than a decade later. Talk of “smart homes” began in earnest around the year 2000 but its US market is only $20 billion today. OLEDs were first used in phones in 2007, yet their revenues were only $46 billion in 2020 and the first VR headset was released in 1991, yet its revenues are only $16 billion.
Big Data and its successor, AI, are particularly disappointing because they are the technologies that were supposed to bring us the productivity improvements necessary for an accelerating digital transformation. Not only are their market sizes still small, but they have been heavily criticized for both impacting low-income and minority groups and for failing to deliver technically. Those concerns indicate that their impact on productivity growth might be even smaller than their market sizes suggest. One of the most prominent early critiques of Big Data and its algorithms came from Cathy O’Neil, whose 2016 book Weapons of Math Destruction described the impact of algorithmic assessments on people’s lives. The algorithms help criminal justice systems determine bail and sentencing based on a person’s associates and neighbors, companies determine employee schedules whose weekly hours are just short of giving them health coverage, and universities game rankings. Most of these applications involve racial and gender bias, worker exploitation, and advertisements that target low-income people with little self-confidence. In a book with hundreds of examples in more than 10 industries, however, we could find few examples of solutions that would enhance productivity.
Why is the slow growth in new technologies in the 2010s a problem? Among other reasons, because their future growth depends on the base that was established for them in the 2010s, and this base remains surprisingly weak. E-commerce, enterprise software, online advertising, social networking, and cloud computing now have huge markets because large bases for them were established decades ago in the 1990s and the 2000s, and as a result, exponential growth has built from this large base. But without a strong base of growth in the 2010s, it is unlikely the newest technologies will achieve high market sizes by 2030, and thus they cannot have a big impact on productivity by then. This is simply how exponential growth works.
Explaining why the growth in new technologies was slow during the 2010s is a more complex and challenging question, one that must be left to other articles. The point we wish to make here is that digital transformation has not achieved what many had expected by 2020 and thus productivity improvements will likely take much longer to emerge than techno-optimists have thought they would take. Comparing market sizes between newer and older technologies adds to the pile of evidence suggesting that our current “tech” industry is not what its boosters claim it is. If we want higher productivity growth, we need to develop different ways of thinking, but first and most of all we must come to grips with the bleak realities of our era, with all of its big promises that simply haven’t delivered.