The manufacture of steam-engine time

The myth of progress and the dangers of monoculture

The manufacture of steam-engine time

The myth of progress and the dangers of monoculture

Photo by Fabrizio Verrecchia on Unsplash

Technological determinism presents a tempting narrative. We see society shaped by technology and technology shaped in turn by society, and we feel justified, by that tight coupling, in saying that technologies that failed did so for a good reason and that our current lineup, because it survived in the hyper-speed pseudo-Darwinism of the global market, is fitter in some general, eternal sense. Path dependence, because it’s so difficult to predict ahead of time, gives us the illusion of destiny. The phenomenon of simultaneous discovery bolsters this feeling — “with steam engines comes steam-engine time” because until that point the prerequisites for steam engines were not in place, and come “telephone time” six people file patents simultaneously.

It is, however, an illusion. Technology (as the class of all human-made items) doesn’t “want” anything, although specific technologies have particular prerequisites and particular side effects that make them fit better in particular environments. Unlike the pseudo-Darwinism of global markets, this is a real Darwinism: the only metric for success is existence, and fitness is always contingent upon the environment. A technology that is fit for 1998 is not fit for 2018, but likewise, a technology that is fit for 2018 is not fit for 1998.

Path dependence is real, but is also mere convenience writ large. We use the materials that are available, and the ideas that are available, and cater to available incentives. Path dependence locks us into a route, and we value that route because of what we have seen at the end of it, despite being totally ignorant of what lies along diverging routes. As our technological path influences our social path, we move toward a future based on present convenience, with no reason to believe that it’s the best possible future (or even any reason not to believe it is the worst one).

Photo by Caleb Jones on Unsplash

Simultaneous discovery is a symptom of a similar kind of mental imprisonment. We produce ideas via a random walk through our inner semantic landscape — we move between similar ideas. Occasionally, we will make a connection that is novel enough to be patentable or protected by some other IP law — in other words, the state, not knowing how to increase the number of novel ideas, merely incentivizes ideas that rise above a particular threshold of novelty. If a connection is novel enough to be patentable, that really just means that the person filing the patent has a mental landscape different enough from the patent examiner that different connections seem obvious, but similar enough that these new connections can be communicated. (The relative nature of novelty is highlighted by the domain of software patents, wherein basic concepts or forty-year-old practices get approved by patent clerks whose background is generally in mechanical engineering.)

Ultimately, simultaneous discovery means that the people working in a field are part of a monoculture — enough people have similar enough backgrounds that they produce the same thoughts at the same time. Having a monoculture like this is a waste of effort: these people could be producing different ideas of equivalent value, if only they had been diversifying their reading material!

Simultaneous discovery doesn’t mean a global monoculture — more often, it means that there’s a specialist monoculture. Everybody with interest in a particular field & access to the means of production has similar life experiences & is reading the same canon of books — the kind of isolated culture initially made possible by broadcast media and currently maintained by the amplification of broadcast-seeded microcultures by social media stovepiping. As we’ve seen in Silicon Valley, this kind of homogeneity produces inventions that are a poor fit for everyone outside the bubble. (And, of course, the now-narrowed path is narrowed further by market forces, which eliminate possible technological lineages based on their ability to make money, often by impressing an even smaller and more isolated group of investors. The association between this and utility in the wider world is approximately zero.)

If we want our technological paths to work for us, we will need to tame them — to perform technological husbandry.

Photo by Martin Adams on Unsplash

The first step is to lower the forces of path dependence by going out of our way to reward seemingly-pointless, seemingly-impractical work for its novelty — in other words, to cause each current path to bloom into thousands of branches. The second step is to cull paths based not on market forces or appeal to a group of rich eccentrics but based on the possibility of long-term value to society. (We need to cull these paths late — let them develop unguided for years — in order to avoid making short-sighted decisions about their value.) The third step is to make sure all of the work is documented and available to the world. This produces a proliferation of paths, and provides the material with which to impress upon people the variety of possibilities.

Once we have created a cabinet of sufficiently distinct possible futures, we should work to disrupt forces that encourage monoculture. This is not a matter of individual improvement: systemic problems require systematic solutions. One way to discourage monoculture is to discourage both broadcast communication and power-law orientation: direct people toward projects that are understaffed, rather than ones that are currently succeeding.

This kind of project does not require government integration. In fact, I would discourage it as a part of any large institution. Taking as a model small writing awards and chindogu competitions, I recommend the formation of many fully-independent small annual competitions for the seeds of novel ideas, each providing a small number of bigger prizes for former entries that have been developed in interesting directions five to ten years later.