Michell Zappa has a fascinating infographic attempting to lay out timelines for future technologies over the next 25 years. It’s an impressive job of collecting data and laying it out in a way that someone can explore. It’s worth playing with. Click through on the link and you can zoom in and drag the graphic around to see what he’s projected, based upon predictions he’s collected from a dozen or so thinkers.
I do wish the infographic were more of a starting point for exploration. I want to click on some of the circles depicting future technologies and see what he’s using as a basis for the projection.
There are some things missing from the graphic as well, and some things on there that I think are implausible.
In general, when we think about what’s going to come down the pipe in technology, it behooves us to think about economics. What are the costs and cost trends of various technologies (either in R&D or development) vs. the demand for them or their economic return?
For this reason I think his projections about space (a lunar outpost and a space elevator both around 2030, for instance) are either implausible or will happen at a small scale. True, NASA has announced plans for a lunar outpost, but it appears to be backing away from them. Such an endeavor would, after all, be incredibly expensive, and offers little in the way of economic return. A space elevator, on the other hand, would lower the cost of access to space, but its guestimated $1 Trillion cost puts it out of the range of capital outlays any country or set of countries will consider for the coming decades.
On the other hand, I think Zappa under represents the impact of biotechnology and energy efforts over the coming years. An aging population creates a nearly insatiable consumer demand for new and better medical treatments. A growing and increasingly affluent population creates a tremendous demand for more agricultural output (especially as people move increasingly to eating meat, which requires far more land per calorie) and for more energy. The combination of apparently stagnating worldwide oil output and the eventual realization by most of the planet that we need to tackle climate change will force us to make the increasing energy supply a greener one.
Zappa covers Green Energy in the chart a bit, but it doesn’t quite convey that solar photovoltaic electricity, for example, will likely drop below the cost of coal electricity by 2020 (if not earlier), or the likely importance of biofuels created by genetically engineered organisms as ‘drop-in’ replacements for gasoline and kerosene. These are both quite near term impacts that will have a larger impact on the planet than space technologies, robotics, or artificial intelligence.
Zappa has a “Biotech” node which includes both a bit of medicine and a bit of food, but again I would have loved to see a more quantitative approach. Between now and 2050, population will rise by 35% and food demand will rise by 70-100%. Arable land, on the other hand, will not increase. One of the prime applications of biotech (in a broad sense) will have to be the increase of food yields per acre to meet that increased demand. Similarly, as the population ages, there will be more and more demand for therapies against the indignities of age. Stem cell treatments are a fantastic advance, but the major killers will remain heart disease and cancer. And the elderly will pump more and more dollars into products that allow them to age more gracefully, helping them look, act, and feel younger.
I very much applaud Zappa for attempting to lay out such a broad view of the future in a single place. It’s a challenging task that required synthesis of information from a number of sources.
Some things, though, are best looked at quantitatively. I would suggest we all think of the future trajectories of technology as being driven by the intersection of consumer demand (how much money are consumers willing to pump into something) and underlying cost trajectories (is the cost of something dropping or rising, and if so how fast?).
In that vein, here are what I see as the biggest trends going forward (with apologies for placing this only in text).
GROWING CONSUMER DEMANDS
1. Health & Longevity:
Worldwide healthcare spending is almost $4 Trillion, or around 10% of the GDP of developed countries, and is growing faster than any other category. Today there are less than 1 Billion people on the planet aged 60 or over. By 2050 there will be 2 Billion.
That doubling of elderly populations in OECD countries and China will place increasing demand for therapies that either slow aging (best case and not guaranteed) or which address the illnesses and loss of function, vitality, and appearance with aging. All up, healthcare will likely more than quadruple in total spending and double as a fraction of worldwide GDP by 2050, to more than 25% of the economies of developed countries and as high as 40% in some.
For instance, the US Council of Economic Advisors predicts that, on current trend, by 2040, healthcare spending will consume 35% of the economy of the United States.
That projected level of spending is more or less untenable. The aging population and rising health care costs will create enormous pressures for new technologies that can address medical needs at lower costs. That is one almost certain prediction.
Worldwide food spending is around $3 Trillion. While worldwide population will increase only by 35% by 2050, growing affluence will lead to a growing demand for less efficient meat and dairy foods.
As a result, total demand for grain (which is needed in large quantities to produce meat) is expected to nearly double by 2050.
At the same time, there is virtually no additional arable land to expand farming into. For farmers to keep up with demand, yields per acre will need to nearly double in the next few decades. At the same time, overpumping of aquifers, debates about pesticides and GMOs, and the energy inputs required in modern agriculture all serve as brakes on productivity gains.
3. Energy & Climate:
Worldwide energy demand is at $4 Trillion today and will roughly double by 2050. At the same time, populations and governments will come around to the need to virtually reduce net carbon emissions. That will place tremendous demand on both low-carbon energy sources and technologies to capture and sequester CO2 and other greenhouse gasses from the atmosphere.
Here is the US Energy Information Administration’s projections for world energy consumption through 2035.
4. Information Technology:
Information technology (including telecoms, computing hardware, software, and online services) is over $3 Trillion today and is growing at roughly 6% per year. Because it is so useful in so many different arenas of life, demand for it will continue to grow. While demand growth will likely slow over time, it could easily be three times as large in 2050, growing past all sectors except Health Care (which will almost certainly be the largest economic sector on the planet by a healthy margin).
Weighed against these consumer demands we have the underlying price and productivity trends.
1. Moore’s Law
Moore’s Law and its analogues for storage and bandwidth will, if they remain on their current paths, reduce the cost of a unit of computation, data storage, and data transmission by an estimated 100 million times by 2050. It remains to be seen whether these trends actually continue. Both physical challenges and potential saturation of consumer demand loom in the decades ahead. If they do continue, for many current applications, storage, bandwidth, and computation will be effectively free. (There will be exceptions for truly massive scale problems in physics, chemistry, biology, neuroscience, and artificial intelligence, where systems are incredibly complex and problems often scale extremely sub-linearly. These areas may be the prime economic drivers of continued improvement of IT power / $$ by mid century.)
2. The Dropping Cost of Genetic Information Processing
..will have a profound effect on biotechnology and medicine. We are much farther from personalized medicine than Zappa’s graphic would lead one to believe. But the Moore’s Law-like exponential drop in the cost of gene sequencing and gene printing will reduce the cost of sequencing a whole human genome to $5 by 2020 and pennies in 2030. In fact, the price of sequencing genes and of printing gene sequences has been dropping far faster than Moore’s Law:
The resulting flood of data, combined with the continued exponential rise in computing power, will start to make possible the large scale data mining necessary to truly extract valuable medical insights from the genome. Cheaper gene printing, cheaper proteomics, and cheaper experimentation systems based on similar trends will start to make an impact on delivering therapeutics, and also in turning manufacturing via synthetic biology into reality.
3. The Exponential Drop in Green Energy Price/Performance and Density
…if it continues, will herald a green energy revolution. Humanity’s energy use, from all sources, is roughly 1 / 6000th the amount of energy that the sun delivers to the planet. It is a huge and largely untapped resource with practically no greenhouse gas emissions. Solar power up until now has been uneconomical due to low efficiencies and high manufacturing costs of solar technologies. But over the last 30 years, solar photovoltaic cells have increased in energy returned per dollar of manufacturing cost by around 7.5% per year. On current pace, they will cross the price of coal-powered electricity between 2015 and 2020, and be half the price of coal around 2025-2030.
Solar does not solve all problems, of course. Intermittent power supplies (due to nights and cloudy days) make it an imperfect solution, but advances in energy storage will allow solar stations or home solar systems to store up energy during sunny periods and return it during the rest. By the 2030s solar should be fully competitive with coal (the cheapest fossil fuel energy source) for most of the world.
The predictions above could be wrong, of course. The world is full of surprises. Trends in consumer demand sometimes change. Exponential trends in technology are even more suspect and more likely to eventually flatten out. But whether these projections are right or wrong, if we want to have a rigorous look at the future, we should attempt to do so quantitatively, putting together our best data on what people want, on what’s possible, and on the trajectories of both.
As for how to sum this up in a single wonderful graphic, I leave that to someone else today. Perhaps Michell Zappa will take a shot.