In this month’s The Atlantic, Matthew Shaer pens an important article, A Reasonable Doubt, about the technology and sociology of DNA matching and its use in jurisprudence. One of the recurring themes we often consider vis a vis Robot Futures is how power relationships are influenced by technological progress, and there are multiple layers to consider in the case of DNA testing, all borne out by this article. First, there is the initial presumption of infallibility many of us have because of our pro-technology bias. Of course we discover that the black and white is in fact grey (thanks Stephen Crane) and that the technology of DNA matching, given a mixed sample, is so incredibly qualitative that ten labs can provide ten answers. Depressing, but eye-opening. But then there is the forward technology march: let’s solve the problem by removing humans from the equation! Shaer goes on to describe a Pittsburgh startup that uses a trade-secret protected algorithm to match DNA samples automatically, without depending on human judgement. Much as war-fighting robots are supposed to avoid human ethical shortcoming by avoiding human decision-making, so this DNA matcher is seen in a dozen states as the new gold standard because there are fewer places for a lab worker to be directly involved. But, ironically, the algorithm is hidden behind the veil of a trade secret, and so our ability to truly audit the process goes missing, and our faith in technology is only further amped. We need a new field, a Sociology of Technology field, similar to VTSS but more short-term; we need to understand just how we identify shortcomings in technology-human systems, and then how our techno-optimist solutions often drive the invention of yet newer problems. Cognitive Tutors, AI assistants, Automated DNA matching– all of them are so much more nuanced in systems-level effects that we make them out to be.