As with most times, it is presently a very interesting time to be a scientist. Moore’s law has begun to falter and most of the attainable victories in microchip fabrication seem to be behind us. Single core processor speeds are simply not growing at the rates they once did. A good PC from years ago may still constitute a good PC today. Absent improvements in raw computing power, we’ve instead seen cost, thermal efficiency, and energy efficiency leading the way in competition. The smart phone industry has been a driving force behind this, commodifying unusually small, robust, passively cooled, and moisture resistant chipsets that bring a degree of portability and disposability to computation never before seen.
Such developments present an interesting existential challenge to all fields of research. The hardware and software technologies at play are now robust enough that no defensible excuse remains for not adopting computational methods. However, as gains in processing performance are now being driven by increasingly parallel chip designs, the onus is on the researcher to learn how to make use of such technologies. There is still room to excel, the technology is no guarantee, and the easy gains in past research methods due to inevitably increasing horsepower may not be expected to continue indefinitely via current technologies.
This point is largely driven in Richard A. Jorgensen‘s 2011 article, “We’re all computational biologists now…next stop, the global brain? ” In this article, Jorgensen recounts the initial divides between molecular biologists and mainstream biologists following the advent of modern molecular techniques. Notably, as genetics was steeped in statistical, molecular, and computational methodologies from its earliest days, such labels are seldom prepended to the title of a geneticist. Jorgensen asserts a popular belief that most attainable discoveries have already been made, and the increasing transparency of the complexity of our world necessitates techniques beyond bare human intuition.
In a sense this may be true, we live in a world of relative safety and luxury largely provided for by the labors of the past. Still, experience shows we do little to afford ourselves said luxury. In 1930, economist John Maynard Keynes famously predicted that with novel developments in automation technologies, humans would soon work only 15 hours a week to have their basic needs met . Instead of this life of leisure, we simply saw explosive growth in administrative labor and bureaucracies. The The question lingers that when we face our next technological revolution, will we even know what to do with it, and would we even know how?
We can solve more complex disciplinary problems than at any other time in history, but we still struggle to make people care. Anyone can find and comprehend photos of the Cuyahogo river burning, the lessons of environmental purity were soundly delivered in the past. Meanwhile, our methods have become so granular and intensive that reproducibility has collapsed across fields. Metrics may be derived to prove most any biased claim in a manner sufficiently compelling to sell a product yet prohibitively expensive to disprove. I question the utility of computational work as an end all to achieving research goals. Surely within information intensive fields it may be the only viable way to do work, let alone compete. Still, let us look at the case of the pot in pot refrigerator. In the 1990’s, Mohammed Bah Abba invented a novel method of refrigeration without the aid of electricity, permafrost, or cool mountain waters, the pot-in-pot refrigerator . This device used the evapotransipiration of water in a sand layer between two porous clay pots to achieve refrigeration in arid climates. The device allowed locals to keep produce fresh longer without need for sun drying or smoking, enabling rural communities to reduce food poisoning, buffer short term surpluses, and bring products to market intact. This device was simple, it used components that have been available and affordable for the duration of human history, its construction could easily be taught at any age, and it had been invented many times before.
To sum up these tangents, what the pot in pot refrigerator teaches me is this: there may still be attainable gains out there, low hanging, easily communicated fruit to be picked. But, the challenge is in communication, both knowing what knowledge is useful and how best to convey it. Increasingly granular analyses and computations may allow increasingly granular findings, but their inherent complexity allows for the the masking of errors and bias. Tended by the unwary we may rapidly slide towards a wealth of interesting facts and publications, yet lack the societal cachet for applications. We may all be computational researchers now, and as we push against exciting new frontiers, we must be mindful to ensure our gains remain defensible.
- Jorgensen, R. A. (2011). We’re all computational biologists now… Next stop, the global brain?. Frontiers in genetics, 2.
- Zürcher, Boris. “The End of Work.” Economic Ideas You Should Forget. Springer International Publishing, 2017. 163-164.
- Soin, Kanwaljit. “The Art of Pottery in Nigeria”. UWEC.