The Singular Path to Limitless Science
How science will be conducted in the future.
The supplanting of humanity has been portended since automatons, were first conceived by the ancients. For giving fire (technology) to humanity, Prometheus was condemned to eternal torture by a bronze-cast eagle — a task too wearisome even for mighty Zeus. From fire onwards, the primary impetus of scientific enterprise has been the alleviation of the burden of labour in civilisation. The oxen cart was once a marvel of human ingenuity and mastery. Modern science, with its unrelenting momentum, predicts the day when we will have intelligent, sentient machines amongst our midst, performing not only the banal tasks of life, but living themselves.
The Scientific Method
The intrinsic process of the scientific method is to use every incremental prior incremental advance to enable the next. In a post-Renaissance era, the compendium of human knowledge has exploded beyond individual comprehension with sub-specialties ad hoc and ad infinitum. Not only is there just so much to know, but every new study builds on previous knowledge and is likely to be more efficient at creating new knowledge than previous studies, as time progresses. We effectively have at hand an accelerating pace of advancement.
What is astonishing is that the pace of all this advancement itself is accelerating. In all aspects of science and technology, technological breakthroughs are best fitted to the exponential curve. Moore’s Law of the doubling of price-performance of a unit of computational capacity every two years is only one well-known example of this exponential pace. In fact, most technological paradigms are being driven faster than an exponential doubling of efficiency every two years. For example, according to the National Institutes of Health, USA, the cost of sequencing a single base pair of DNA fell by a factor of 10,000 over the last 10 years, which itself is staggeringly 10-times the rate predicted by Moore’s Law.
We are approaching the phase of the trajectory which, in mathematical terms, is called an asymptotic ‘singularity’ — where the rate of scientific and technological advancement is so astonishingly fast that a few days of knowledge creation in this phase outcompetes, in scope and quantity, all preceding knowledge our civilisation will have known. This pace of advancement is truly difficult to comprehend. As is the nature of exponential functions, progress appears linear and slow at first, until reaching a curve from where it shoots steeply and indefinitely upward. The real question at heart is: when will we experience this sudden curve, followed by the singularity of the exponential trend, and what does it mean for us in practical terms?
Scientist of the Gaps
Due to the sheer volume of knowledge that exists and is created every day, scientists increasingly struggle to to keep pace in their area of sub-specialisations. Though focus is a requisite for pushing boundaries, it can lead to a myopic vision vis-à-vis science-at-large and its future. After thinking of an interesting experiment, the scientist’s task is to simply follow logical steps — working as a conduit between different pieces of machine-equipment in order to collect precious data. For example, molecular biologists take blood samples, separate and prep certain cells using predefined methods and kits, finally feeding those cell contents to specialised, preprogrammed machines. Increasingly, the form of data acquired from experiments is digital. It is safe to say that through technology, we have already offloaded and outsourced the development of science and technology itself.
Before we approach the asymptote of the singularity, the single obstacle limiting the productive capability of human civilisation is the human aspect. Automated machines in a lab, networked with each other, could work relentlessly, increasing productivity far beyond human-driven ranges. With all of the time-consuming, mundane tasks relegated to the machines, might scientists finally focus their curiosity and creativity on conceiving experiments? Advances in robotics are accompanied by advances in artificial intelligence (AI), meaning that the same machines that will carry out an experiment may also be better at conceiving and analysing them. Artificial intelligence and robotics have only just started to penetrate the scientific world. According to futurist and current head of engineering at Google, Ray Kurzweil, the year 2029 will see the birth of computers as ‘intelligent’ as humans. By this time, the trajectory is that laboratories will have been almost completely automated. Machine precision, coupled with the maturation of lab-on-chip techniques, will usher a revolution in scientific data mining, sample handling, throughput, statistical analysis and modelling.
After 2029, AI will dream of the next big experiment, will design the experiment, perform it to sheer precision, obtain results (that it may have already predicted), use them to further the models of our then-known universe, and repeat… ad infinitum. Machines will displace and indeed assimilate humanity, both in labour and agency. And this will suit them/us, because what else is there to do except to explore, when faced with the awe of existence? As the celebrated computer scientist Marvin Minsky put it, ‘Will robots inherit the earth? Yes, but they will be our children.’