Peak Tech Is a Tale of Evolution and Extinction
(Bloomberg Opinion) -- The history of technology is one of emergence, finding markets and eventually being challenged by other, newer platforms. New technologies can be tracked along an S-curve, starting from zero, growing slowly, then quickly, before hitting a natural plateau. These S-curves emerge in many different sectors, at both industrial and consumer scale. For an elegant illustration, we can look to the 90-year history of broadcasting licenses for radio and television in the U.K.
Each curve is steeper than the one before it. Only radio has a noticeable decline on the way to its peak, during World War II. For long-term planners in the energy sector who must consider the growth of particular technologies, the S-curve suggests a predictable path to peak deployment. What happens after that, though, is less predictable.
Electrical systems analysis, in particular, assumes that technologies are deployed based on the cost and efficiency of different types of power plants, the costs of fuel, the age of assets, and the needs of operating an electrical grid. A feature of deployment in electricity is that even if technologies are no longer new, they remain part of the system: For example, in the U.S., we use much less oil to generate power than we did 50 years ago, but we don’t use zero. There are century-old hydroelectric plants still generating power today, though no utility or power producer is building new ones.
There are two other features of large power plants that lend credence to the assumption that electrical-generation technologies eventually reach some sort of equilibrium. The first is that they’re big — physically large, centralized, planned over a long time, cost hundreds of millions or billions of dollars, and operate in a stable policy and regulatory framework.
The second is that technologies aren’t the same, even if they meet the same ultimate goal of delivering electrons; some technologies, such as nuclear power, are designed to run at peak capacity indefinitely (except for maintenance), while others, such as reciprocating engines burning natural gas, are only meant to run at designated times. Every technology has its place in a properly functioning system.
Let’s return, then, to those licensing S-curves. With the introduction of and licensing of new broadcast technologies in the U.K., incumbent technology licensing declined. But it didn’t just decline: It vanished.
After each technology’s S-curve peaked, it proceeded to disappear. There’s no equilibrium between licenses for black-and-white TV and color TV. Plot these technologies another way, and technology deployment is less a collection of overlapping curves and more a series of waves wiping out what came before.
The total market for these technologies has stayed proportional to the number of households in the U.K., but the composition of that market has changed significantly as one technology replaces what came before it.
Technology analyst Horace Dediu wrote about the rise and decline of technologies to their equilibrium in a 2012 post Of BlackBerry and the “language commonly heard among analysts,” he wrote:
The company was "reverting to the mean" and growing nearly in-line with the market. In other words, exceptional growth was over but continuing growth was likely. The company was returning to something “normal.”
But, as with U.K. broadcast technologies, there was no mean reversion for BlackBerry. Instead, there was only “an acceleration toward oblivion”:
Disruption happens. It happens quickly and perhaps most quickly in this industry, and looking at trailing indicators like revenues and profits is of no help. I’ve only witnessed sharp expansions and sharp declines. I’ve never witnessed steady state, quiescent markets in technology. The history of platforms is always virtuous cycles followed by vicious declines.
The examples of BlackBerry and TV licenses give us a number of things to consider for other technologies.
The first is that new power technologies — wind, solar, batteries — are highly distributed and therefore have a different total market than coal and nuclear plants. A dramatically cheaper coal plant might not get built as the grid might not need its output; dramatically cheaper solar creates its own market, as do dramatically cheaper batteries. At the same time, that market might be limited to cohorts of the population, such as those who own their homes and are willing to purchase an asset that will last for the term of a mortgage or longer.
Or it might not be. In Australia, solar has already reached 6 percent penetration of rental properties as investors purchase solar-equipped homes and rent them out. As that number rises, we will need to revisit our assumptions about which types of buildings will host solar power and therefore reassess Australia’s total market for solar power.
The second thing to consider is that rapidly improving, highly distributed energy technologies can leapfrog their logical predecessors; think mobile phones replacing landlines in developing countries.
Electric vehicles are an example, too: Automakers' tech roadmaps tend to move from pure combustion vehicles to hybrid vehicles to plug-in hybrid models that have both a battery and an internal combustion engine to all-electric vehicles. There are lots of assumptions within those roadmaps, including that consumers would want vehicles with both batteries and an engine for reasons of cost and vehicle range. That’s not what's happening right now: As batteries improve and become cheaper, and as consumers' preference for pure electric vehicles takes shape, automakers are jumping right past plug-in hybrids and directly to pure electric cars.
Look at a technology’s S-curve in isolation, and it shows growth to a point of saturation. That saturation, though, isn’t happening in isolation; it is displacing some other technology, sometimes quite dramatically. Tech platforms are either growing or shrinking; they don’t necessarily achieve equilibrium. Each wave of growth doesn’t necessarily spare what came before it; sometimes one wave wipes out its predecessor before it even begins to develop. Something to think about for system planners.
- The Bloomberg 50 people with noteworthy accomplishments in 2018.
- The U.S. Alliance for Sustainable Finance convenes 15 founding financial institutions “to drive investment in clean energy and resilience projects.”
- Russian and Saudi Arabian cooperation in the oil market will not last.
- Luxembourg will make all public transportation free starting in 2020.
- Lyft has (confidentially) filed for a U.S. initial public offering.
- Uber doesn’t have to follow Pan Am’s tragic arc.
- As Seattle approaches what its traffic engineers call the “period of maximum constraint” UPS has begun delivering packages via electric tricycle.
- Half of Americans think food additives and genetically modified organisms are unhealthy. The other half isn’t concerned about them.
- There are 720,000 people skilled in machine learning in the global workforce.
- An “anatomical map” of an artificial intelligence system in the form of an Amazon Echo device is “an exploded view of a planetary system across three stages of birth, life, and death.”
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Nathaniel Bullard is an energy analyst, covering technology and business model innovation and system-wide resource transitions.
Hugh Bromley leads BNEF’s analysis of North American solar and distributed energy markets. An accredited photovoltaic engineer, he has over five years of experience in solar and power markets across North America and Asia-Pacific.
©2018 Bloomberg L.P.