Have Technological Advances Peaked?
Have Technological Advances Peaked?
Complete the form below and we will email you a PDF version of "Have Technological Advances Peaked?"
The seemingly high rate of technological growth is illusory: the world is slowing down and will continue to do so long into the future. HSE scholars predict that the “technological singularity” will occur in 2106 and that, contrary to what some expect, it will not mark the apotheosis of progress. Here, Andrey Korotayev, Leading Research Fellow, Head of the HSE Laboratory for Monitoring the Risks of Socio-Political Destabilization and a co-author of this study explains why global acceleration is now a thing of the past, prognosticators are unafraid of “black swans” and the coronavirus will not rewrite human history.
How did the idea originate of a “singularity” — a point at which technological growth becomes uncontrollable and irreversible — and why do you disagree with it?
For many millennia, technological development advanced along a hyperbolic curve of constant acceleration. Ultimately, this curve seems to accelerate into infinity. This is best described by a hyperbolic function with an exit point at infinity—that is, a singularity. Taken literally, at this point, the constantly accelerating progress will become uncontrollable by man and fall under the control of artificial intelligence.
I think this understanding is wrong. Acceleration naturally gives way to deceleration. Global population dynamics provide a clear example of this mechanism. In 1960, Austrian scientist Heinz von Foerster showed that the number of people living on the planet has grown according to a hyperbolic progression. Since the 1970s, however, this trajectory has transformed — the growth rate has begun to decline. This is still the trend, and the UN predicts that it will continue for decades to come.
We observe a similar pattern concerning technological progress, which is always linked to global population dynamics.
Therefore, the technological singularity is not a point at which the curve shifts into the infinite, but at which the type of development changes — from the hyperbolic acceleration seen throughout almost all of human history, to a slowing of that growth rate.
What will happen in 2106?
Judging by the current data and empirical evidence, a singularity already occurred in 2018 and we now live in a decelerating world. Another wave of acceleration awaits in 2030, however, the point of singularity for which will come in 2106 — give or take 20-30 years because such predictions are estimates. In any case, the pattern of technological development will change in the 21st century.
On what do you base your prediction?
We can anticipate such transformations by identifying historical patterns. For this, we apply principles of the theory of production — that is, long, overarching cycles during which innovations appear and technological changes of critical importance take place.
Human history has seen four such cycles: the hunter-gatherer, agrarian-trade, commercial-industrial, and cyber. Because of acceleration, each cycle has been shorter than the last.
The hunter-gatherer period lasted 30,000 years, the agrarian-trade 9,400 years, the commercial-industrial 525, and the cyber began in 1955 and will last 135 years-160 years.
Nonetheless, each cycle follows the same basic pattern: each contains six similar phases whose sequence and duration remain stable even as the principles of production changed.
These constant ratios make it possible to calculate the dynamics of technological progress since the time of the Upper Paleolithic Revolution of 40,000 BC and to predict its future stages of development.
What is the current stage?
We have been in the second stage of the Cyber Revolution since the mid-1990s. It began in the 1950s with the rapid growth of information technologies, energy production, automation, space exploration, the transition to scientific methods of managing production, and the globalization of the economy. With the third stage that will come in approximately 2030, the final phase begins that could become the era of “smart” self-regulating systems. Its driving force will be the MANBRIC convergence, the synergy produced from the development of medicine (M), adaptive technologies (A), nanotechnology (N), biotechnology (B), robotics (R), the IT sector (I) and cognitive technologies (C). With the fourth phase that will start in 2055, self-regulating systems will begin improving rapidly and will spread around the world at incredible speed, attaining highly developed forms and occupying a central role in the new production process.
For good reason, medicine will be at the forefront of this movement. The structural change inherent in the ageing of the global population will drive technological growth.
The number of 80-year-olds on the planet between 1950 and 2050 will increase 100-fold, with a corresponding increase in the number of people spending money on health. This will provide a powerful incentive for the appearance and commercialization of breakthrough solutions.
But if that growth will be so significant, why does it amount to only a wave before a slowdown?
I see no reason to assume that the acceleration of the 2030s will be greater than that of the 1950s-1960s. There were breakthroughs everywhere then — in transport, energy, the chemical industry, molecular biology and so on. That laid the foundation for the appearance of personal computers in the 1980s
A similar boom is unlikely. The acceleration with continue, but only in a few isolated areas.
In artificial intelligence, for example, experts anticipate breakthroughs in only two or three of 12 different fields, with growth slowing in the rest.
One of the indicators of the state of technical progress is the fact that the number of breakthrough inventions had already begun to slow by the end of the last century. Statistics indicate that a great many patents have been filed, but many of them are trash: patents are often granted for things that are not full-fledged inventions. The number of truly important patents has declined since the 1950s-1960s.
Is the human race exhausting its potential?
This is a matter of base causes. The idea that technology is constantly accelerating comes from the old understandings of the universe as endless in time and space. Now we know that the universe is endless but, of course, it has a definite beginning and a limited amount of matter. The number of protons, neutrons, and electrons has been more or less calculated. The hope turned out to be groundless that by studying their structure we would discover new elemental particles and move forward. This necessarily limits the number of discoveries possible.
Thus, fundamental physics reached its peak of discoveries in the 1920s-1930s. The field commercialized and began producing economic benefits in the 1950s-1960s, but the situation became less promising with time.
We can confidently state that the 22nd century will see significantly fewer discoveries than the 21st, and the 23rd fewer still.
World-changing inventions will continue to appear, as will periods of accelerated technological growth lasting decades. Of course, the 2030s will be more productive than the 2010s, but the overall trend will be towards deceleration.
We need to prepare for a life of slowing economic growth and overcome the idea that this is temporary. The negative dynamic that developed countries have experienced since the 1970s is not the result of wrong economic policy. It is a deep and serious trend. The technological boom of the 2030s will improve the dynamic of the global economy, but will not reach the level of the 2000s, much less the highpoint reached in the mid-20th century.
Could the appearance of such “black swans” as the coronavirus and the resultant economic crisis force you to reconsider these predictions?
The coronavirus pandemic fits well into the long-term forecast. Consider an earthquake. Experts would doubtless tell you that Los Angeles is located on a seismically active zone and that skyscrapers there must be built with reserves of strength and stability. But nobody can predict the timing of the aftershocks. The same is true of epidemics. The model suggests with high probability that they are inevitable, but it cannot predict exactly when or with what force they will strike. Predictions also exist for a global economic crisis. We have known since the 19th century that an economic cycle — that is, the fluctuation between boom and recession — last for from seven to 11 years. The most recent crises occurred in 2008 and 2014, and so a new crisis was expected in 2020. The world economy was ripe for one, and it would have come this year or the next, regardless of the coronavirus.
Does the pandemic have a “point of singularity”?
There is no point at which the pandemic will shift into endless progression. It will end at some point. Some adaptation will then take place and the consequences in the 2060s and 2070s will be less severe. There is no reason to expect ultra-high mortality rates from such waves of pathogens because the technology for countering them is constantly improving.
Do cataclysms spur more rapid technological progress thanks to discoveries that would not ordinarily occur during less turbulent times?
On the scale of a decade, yes, but on the scale of a century, no. They can trigger new waves. The wave would have come anyway: the discoveries or cataclysms just speed their appearance.
Even the asteroid that struck Earth 60 million years ago and caused the extinction of countless species, or the Black Plague in the 14th century that carried off one-third of Europe’s population, look like nothing more than fluctuations on the long-term curve in the overall trend — a momentary downward deviation before returning to its starting point. These fluctuations can be significant, but the trend does not change.
Grinin et al. (2020). A quantitative analysis of worldwide long-term technology growth: From 40,000 BCE to the early 22nd century. Technological Forecasting and Social Change. DOI: https://doi.org/10.1016/j.techfore.2020.119955
This article has been republished from the following materials. Note: material may have been edited for length and content. For further information, please contact the cited source.