In recent years, we have seen rapid innovation and improvements in different technologies, such as AI, 5G, blockchain technology, AI/VR, and computation and storage. There is great belief that the integration of these technologies will create a new “techno-economic paradigm” and will thus bring large revenues and returns. However, creating a new techno-economic paradigm is never solely a matter of technological innovations as such, but also of developing and expanding the (eco)system in which these subtechnologies are integrated, and history teaches us that this is never a smooth process but comes with bubbles and crashes. As such, we may speculate about a “smart bubble” being blown up around current digital solutions and companies.
- 2019 was a good year for many publicly traded and relatively old tech companies, but not so much foryounger tech firms that went public only recently and those who are still dependent on VC. In fact, one could argue that they experienced a small crash (e.g. WeWork, Uber and Juul). Most of them are businesses built on the belief that digital technology can disrupt any sector, but in practice, these companies have failed to make a profit just yet. We have written before that these Big Tech companies rely heavily on narratives of how digital technology can make our world a better place.
- Previously, we have written about technological revolutions of the past and future. Today, we’rewitnessing the emergence of the next technological revolution, which is characterized by smart technology that will lead to increasingly autonomous systems. The techno-economic paradigm that could result from these developments can be characterized as the sensor-based economy.
- According to Carlota Perez’ scheme of technological revolutions, enthusiasm about new technology and its commercial value (always) leads to a frenzy, whereby far too much capital is poured into technological development and infrastructure build-up. This was true, for instance, for the railroad in the early 19th century and more recently with the dotcom boom. These frenzies necessarily lead to a major crash and only afterwards can a more sensible and rewarding phase start, during which the technology reaches full maturity.
- Investments in AI have grown exponentially over the course of the decade and Big Tech companies are heavily involved through internal R&D and acquisitions of dedicated AI developers. Although quantum technology is years behind AI in terms of practical applications, investment in this space has grown exponentially as well (albeit far less in absolute numbers). Hundreds of billions of dollars will be spent on 5G licenses and infrastructure rollout.
Connecting the dots
As tech stocks have fared extremely well over the last years, analysts are obviously starting to wonder whether this winning streak is coming to end. Yet, few appear to believe that this will happen in the coming year, and if it does happen, it will most likely be due to a general economic downturn during which tech stocks may be hit hardest, simply because the bigger they are, the harder they fall. In 2019, several high-profile tech companieshave failed to meet expectations, e.g. WeWork, but most of these companies are not listed (yet). And, as several observers have mentioned, these are not really tech companies, but rather less fruitful branches of the disrupt-everything tree in sectors in which digitization does not bring any radical cost savings or zero-marginal cost scale effects (e.g. humans still drive the cars, renting out scooters or office space).
Beyond current, by now conventional, digital technology, the next tech revolution is already in the making. As we have noted before, the next revolution is all about artificial technology and autonomous systems, which the WEF has dubbed the Fourth Industrial Revolution. As we have learned from the past, and particularly from Carlota Perez’ analysis, technological revolutions tend to lead to an initial frenzy that ends with the bubble bursting anda lot of capital going to waste. To put matters into perspective, the previous IT revolution enjoyed its (dotcom)frenzy in the 1990s and was temporarily interrupted by the dotcom crash of 2000. The crash, as in Perez’ scheme, was followed by a period of more realistic growth during which technology companies actually satisfied a real need and made real profit (i.e. the deployment phase), until they eventually ran out of steam and made way for the next revolution. On a speculative note, last year’s tech disappointments may even be a sign that these are the last gasps for breath of the old IT revolution and that a new tech revolution is in the making.
Arguably, the unfolding smart revolution (for lack of a better term) is currently enjoying its boom as investments are piling up for technology that is surrounded by great expectations, but has yet shown relatively few real life and value-adding applications. Moreover, while AI, for instance is being used already in “narrow” applications (e.g. image recognition) and the first 5G networks have been built, the true promise of these technologies relates strongly to the larger paradigm they may bring about once they are all in place and start working together (e.g.the WEF’s Fourth Industrial Revolution or the sensor-based economy).
In other words, the true promise of the next revolution is shaped by a stack of promising individual technologies. Cheaper and more capable sensors will produce tons of data, next-gen network technology willallow fast and reliable transmission of this data and new forms of intelligence in the cloud and in the edge will make sense of all raw data and generate new insights and products. All of this may take place in decentralized and trustless networks that are more robust and prevent single actors from becoming too powerful. As such, while promising technologies underpin the hope of the new paradigm on different layers of the stack, each individual technology also presents a risk to the development of the entire stack (which can also be understood as a house of cards). That is, if one or more of the enabling technologies fails to meet expectations “in time”, the entire stack may fail to do so as well.
There are more threats to this prospective stack. Implicitly, the various elements in the stack rely on each otherfor their commercial success. To illustrate, 5G investments can only be recouped when sufficiently popular applications are developed and AI can only become truly valuable when a range of sectors are digitized further so that the AI system can be fed with relevant data and insights can actually be implemented (ideally in real time and in autonomous systems). What’s more, future regulation could hamper (for good reasons) the unfolding of the new paradigm. That is, anti-trust measures would likely break up Big Tech companies that are building integrated and data-driven ecosystems from their various businesses, which mutually reinforce each other, and help these companies to tap into new markets to gain even more data and sources of revenue. Breaking up these “digital conglomerates” in the name of antitrust could create a division of currently concentrated capital (potentially hampering R&D efforts) and the vast pools of data in the hands of Big Tech, the data needed to train AI systems, and make for effective autonomous systems. Furthermore, more stringentregulations on data sharing to protect citizens and consumers could hamper flows of data throughout the prospective stack and (again, for good reasons) limit the potential of future smart systems. It should be noted that, while these are potential headwinds, history has also taught us that large incumbent firms are not necessarily the most innovative companies, hence breaking up Big Tech could be worthwhile, and we’ve also learned that regulation can actually be a powerful trigger for innovation, and in this case could make for a “better” revolution from a societal perspective.
If and when this new bubble will burst is difficult to say, but there are some arguments to take into account.Perez’ timeline suggests it may take another decade or so before this bubble bursts (although it would be difficult to determine the starting point). At the same time, there are also good reasons to claim thattechnological progress is speeding up (if only because tech companies build on each other’s knowledge, tools and infrastructure) and the sheer availability of capital (e.g. due to low interest rates and high free cash flows)further adds to the speculative bubble (as happened with WeWork and others). Looking at the individual technologies, one could expect, for instance, 5G deployment to scale up massively, but it is also likely that critical uses of 5G (e.g. autonomous vehicles or VR applications) are further away and that it will be difficult to recoup investments in 5G technology and infrastructure. It is also questionable whether AI will find meaningful, scalable and legal applications that add enough value to provide a return on current investments.
- If more signs and evidence appear of an impending smart bubble, investors will have to reflect more critically on whether digital solutions are rendering real value and solve real problems. In such a critical analysis, the term “smart” could come to define a much narrower domain of technological solutions and innovations and only include, for example, digital systems that can identify and solve problems themselves, or AI that renders solutions beyond human grasp.
- In Perez’ scheme, the frenzy leads to a “turning point” and to the popping of the bubble. Only during the deployment phase does the technology reach full maturity and generate real added value. From this perspective, one could argue that the next techno-economic paradigm is still in its infancy, and will likely lead to higher economic growth rates and productivity growth in the coming decades, relieving central banks and governments of their economic headaches.