In Lewis Carroll’s Through the Looking Glass, a sequel to Alice in Wonderland, Alice and the Red Queen are running together across the Red Queen’s country. Alice cannot believe how fast they are running and is unsure how much longer she can keep up. Yet no matter how fast they go, they never seem to get anywhere because, as Alice realizes in astonishment, the country they are moving through is moving with them. The Red Queen explains that, in her country, “It takes all the running you can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!”
In evolutionary biology, this sequence has formed the basis for “The Red Queen Hypothesis,” an analogy for the evolvability required of living organisms due to the evolutionary pace of the entire system. In dynamic evolutionary niches, competition from predators, prey, peers, and parasites creates intense evolutionary arms races between organisms wherein standing still means falling behind. It is evolutionary arms races that endow lions, and simultaneously gazelles, with their speed; that underlie the astonishing complexity and adaptability of our immune systems, and the equally astonishing resilience of the pathogens that infect us. In these systems, organisms are pushed to evolve as fast as they can simply to keep up with everything else evolving around them.
Rapidly changing industries and technology areas have similarly intense demands placed on their pace of evolution. All the participants are so intensely focused on innovation that in order to get anywhere, innovators must take the Red Queen’s advice, and pick up the pace. What this means in practice is that innovators must have longer-range vision and must develop the ability to adapt to changing conditions; otherwise, their nearer-term innovations are likely to be leapfrogged by step-change technologies that make more pedestrian improvements obsolete.
The semiconductor industry operates in a state of constant breakneck innovation due largely to “Moore’s Law,” an accepted standard that has demanded doubling of integrated circuit complexity every two years since the 1960s. The companies that make equipment used in semiconductor manufacturing are constantly innovating to keep up with Moore’s law. Their speed of innovation is so rapid, in fact, that often there is not enough time to work out all the kinks with the old system before a new paradigm emerges.
Rapid thermal processing (RTP), for example, developed as a viable alternative to furnace annealing of semiconductor wafers in the late 1990s, and permitted more controlled diffusion and activation of dopants. RTP involves suspension of a single wafer on quartz pins inside a chamber followed by rapid heating by high powered optical lamps. This rapid heating helps control dopant diffusion and allows the formation of shallow transistor junctions, which are essential for enabling sub-micrometer feature sizes. As a result, RTP machines have become standard in all modern chip fabs.
Despite its success, even cutting-edge RTP machines suffer from imperfect heating and cooling control, loss of heat through the quartz pins, inefficient heating of the edge ring of the wafer, and non-ideal dopant diffusion. In 2009, however, a European equipment manufacturer introduced a new and improved type of rapid thermal processing machinery, which they believed would solve these lingering problems. Unlike traditional RTP machines, their equipment heats using conduction, which produces more efficient and even heat transfer, and eliminates the need for quartz pin support. The company’s scientists were solving for very real and clearly articulated problems in the semiconductor industry, yet to date almost none of their machines have been sold to major fabs and the company has since switched its focus to solar cell processing.
So where did things go wrong? It seems that, while they were busy solving the problems of quartz pin support and even heat transfer, the industry jumped years ahead to a revolutionary new technology in thermal processing: laser flash annealing. This method uses a laser to heat just the top few atoms of a wafer in a few microseconds! With this new advancement in thermal processing in play, the large fabs are no longer interested in incremental improvements to yesterday’s technology, and cutting-edge fabs such as Intel, TSMC, and Samsung have now made the transition to laser flash annealing for their highest complexity chips. The European equipment manufacturer’s failure to see the 10- or 15-year plan of thermal processing development has left it floundering without a play in semiconductor manufacturing.
In dynamic industries, nearer-term innovations are likely to be leapfrogged by step-change technologies that make more pedestrian improvements obsolete.
These same sudden, disruptive surges in technical sophistication also occur in other dynamic industries, such as biotechnology. For instance, Genentech’s development, and Eli Lilly’s commercialization, of recombinant synthetic human insulin in the late 1970s revolutionized the treatment of diabetes at the expense of companies producing standard, but suddenly inferior, animal-derived products. Today, production and delivery of human insulin and its standard analogs are being challenged by a similarly revolutionary technology: so-called “smart insulins” that can modulate their activity in response to the levels of glucose they detect in vivo. Large pharmaceutical manufacturers such as Sanofi and Merck have already begun investing heavily in this new step-change technology, which has significant potential to improve compliance and reduce hyperglycemia; if they are successful, slower adapters may find themselves dangerously out of date.
Elsewhere in biotechnology, other recent developments – the CRISPR revolution in genetic manipulation; advances in methods and materials for drug delivery; the enormous potential (today only partially realized) of big data and machine learning to decipher macromolecular structure, function, and interaction – indicate that the pace of innovation is only accelerating. Even less traditionally dynamic industries such as automotive and architecture have seen external pressure from government regulations and increased consumer environmental awareness create an energetic, innovative environment focused on energy efficiency and “green design.” As the pace of change increases and spreads to new niches, an evolvable, long-range innovation culture becomes ever more critical.
In slower changing industries, the tried and true innovation method of identifying an unsolved problem and setting out to solve it is appropriate and clearly successful. In dynamic niches such as semiconductor manufacturing and biotechnology, however, identifying and solving pedestrian problems is a necessary but not a sufficient step. While incremental improvements are important, innovators in these spaces must keep one eye on the horizon. These dynamic industries are firmly within the Red Queen’s country, and here, innovation means thinking longer-term, increasing your evolvability, and running “at least twice as fast as that!”