Where are we going?
From Technological Forecasting and Social Change.
Copyright © 2005 Elsevier Inc. All rights reserved.
Article in Press, Corrected Proof
A possible declining trend for worldwide innovation.
A comparison is made between a model of technology in which the level of technology advances exponentially without limit and a model with an economic limit. The model with an economic limit best fits data obtained from lists of events in the history of science and technology as well as the patent history in the United States. The rate of innovation peaked in the year 1873 and is now rapidly declining. We are at an estimated 85% of the economic limit of technology, and it is projected that we will reach 90% in 2018 and 95% in 2038.
Entering a dark age of innovation
14:00 02 July 2005
NewScientist.com news service
SURFING the web and making free internet phone calls on your Wi-Fi laptop, listening to your iPod on the way home, it often seems that, technologically speaking, we are enjoying a golden age. Human inventiveness is so finely honed, and the globalised technology industries so productive, that there appears to be an invention to cater for every modern whim.
But according to a new analysis, this view couldn't be more wrong: far from being in technological nirvana, we are fast approaching a new dark age. That, at least, is the conclusion of Jonathan Huebner, a physicist working at the Pentagon's Naval Air Warfare Center in China Lake, California. He says the rate of technological innovation reached a peak a century ago and has been declining ever since. And like the lookout on the Titanic who spotted the fateful iceberg, Huebner sees the end of innovation looming dead ahead. His study will be published in Technological Forecasting and Social Change.
It's an unfashionable view. Most futurologists say technology is developing at exponential rates. Moore's law, for example, foresaw chip densities (for which read speed and memory capacity) doubling every 18 months. And the chip makers have lived up to its predictions. Building on this, the less well-known Kurzweil's law says that these faster, smarter chips are leading to even faster growth in the power of computers. Developments in genome sequencing and nanoscale machinery are racing ahead too, and internet connectivity and telecommunications bandwith are growing even faster than computer power, catalysing still further waves of innovation.
But Huebner is confident of his facts. He has long been struck by the fact that promised advances were not appearing as quickly as predicted. "I wondered if there was a reason for this," he says. "Perhaps there is a limit to what technology can achieve."
In an effort to find out, he plotted major innovations and scientific advances over time compared to world population, using the 7200 key innovations listed in a recently published book, The History of Science and Technology (Houghton Mifflin, 2004). The results surprised him.
Rather than growing exponentially, or even keeping pace with population growth, they peaked in 1873 and have been declining ever since (see Graphs). Next, he examined the number of patents granted in the US from 1790 to the present. When he plotted the number of US patents granted per decade divided by the country's population, he found the graph peaked in 1915.
The period between 1873 and 1915 was certainly an innovative one. For instance, it included the major patent-producing years of America's greatest inventor, Thomas Edison (1847-1931). Edison patented more than 1000 inventions, including the incandescent bulb, electricity generation and distribution grids, movie cameras and the phonograph.
Huebner draws some stark lessons from his analysis. The global rate of innovation today, which is running at seven "important technological developments" per billion people per year, matches the rate in 1600. Despite far higher standards of education and massive R&D funding "it is more difficult now for people to develop new technology", Huebner says.
Extrapolating Huebner's global innovation curve just two decades into the future, the innovation rate plummets to medieval levels. "We are approaching the 'dark ages point', when the rate of innovation is the same as it was during the Dark Ages," Huebner says. "We'll reach that in 2024."
But today's much larger population means that the number of innovations per year will still be far higher than in medieval times. "I'm certainly not predicting that the dark ages will reoccur in 2024, if at all," he says. Nevertheless, the point at which an extrapolation of his global innovation curve hits zero suggests we have already made 85 per cent of the technologies that are economically feasible.
But why does he think this has happened? He likens the way technologies develop to a tree. "You have the trunk and major branches, covering major fields like transportation or the generation of energy," he says. "Right now we are filling out the minor branches and twigs and leaves. The major question is, are there any major branches left to discover? My feeling is we've discovered most of the major branches on the tree of technology."
But artificial intelligence expert Ray Kurzweil - who formulated the aforementioned law - thinks Huebner has got it all wrong. "He uses an arbitrary list of about 7000 events that have no basis as a measure of innovation. If one uses arbitrary measures, the results will not be meaningful."
Eric Drexler, who dreamed up some of the key ideas underlying nanotechnology, agrees. "A more direct and detailed way to quantify technology history is to track various capabilities, such as speed of transport, data-channel bandwidth, cost of computation," he says. "Some have followed exponential trends, some have not."
Drexler says nanotechnology alone will smash the barriers Huebner foresees, never mind other branches of technology. It's only a matter of time, he says, before nanoengineers will surpass what cells do, making possible atom-by-atom desktop manufacturing. "Although this result will require many years of research and development, no physical or economic obstacle blocks its achievement," he says. "The resulting advances seem well above the curve that Dr Huebner projects."
At the Acceleration Studies Foundation, a non-profit think tank in San Pedro, California, John Smart examines why technological change is progressing so fast. Looking at the growth of nanotechnology and artificial intelligence, Smart agrees with Kurzweil that we are rocketing toward a technological "singularity" - a point sometime between 2040 and 2080 where change is so blindingly fast that we just can't predict where it will go.
Smart also accepts Huebner's findings, but with a reservation. Innovation may seem to be slowing even as its real pace accelerates, he says, because it's slipping from human hands and so fading from human view. More and more, he says, progress takes place "under the hood" in the form of abstract computing processes. Huebner's analysis misses this entirely.
Take a modern car. "Think of the amount of computation - design, supply chain and process automation - that went into building it," Smart says. "Computations have become so incremental and abstract that we no longer see them as innovations. People are heading for a comfortable cocoon where the machines are doing the work and the innovating," he says. "But we're not measuring that very well."
Huebner disagrees. "It doesn't matter if it is humans or machines that are the source of innovation. If it isn't noticeable to the people who chronicle technological history then it is probably a minor event."
A middle path between Huebner's warning of an imminent end to tech progress, and Kurzweil and Smart's equally imminent encounter with a silicon singularity, has been staked out by Ted Modis, a Swiss physicist and futurologist.
Modis agrees with Huebner that an exponential rate of change cannot be sustained and his findings, like Huebner's, suggest that technological change will not increase forever. But rather than expecting innovation to plummet, Modis foresees a long, slow decline that mirrors technology's climb.
At the peak
"I see the world being presently at the peak of its rate of change and that there is ahead of us as much change as there is behind us," Modis says. "I don't subscribe to the continually exponential rate of growth, nor to an imminent drying up of innovation."
So who is right? The high-tech gurus who predict exponentially increasing change up to and through a blinding event horizon? Huebner, who foresees a looming collision with technology's limits? Or Modis, who expects a long, slow decline?
The impasse has parallels with cosmology during much of the 20th century, when theorists debated endlessly whether the universe would keep expanding, creep toward a steady state, or collapse. It took new and better measurements to break the log jam, leading to the surprising discovery that the rate of expansion is actually accelerating.
Perhaps it is significant that all the mutually exclusive techno-projections focus on exponential technological growth. Innovation theorist Ilkka Tuomi at the Institute for Prospective Technological Studies in Seville, Spain, says: "Exponential growth is very uncommon in the real world. It usually ends when it starts to matter." And it looks like it is starting to matter.