Slate reports that the first news report on yesterday’s L.A. earthquake was generated by an algorithm called Quakebot. A reporter named Ken Schwencke developed the algorithm to extract key data from US Geological Survey reports and plug them into a template he created. He can then just hit “publish.”
Given the seemingly meager amount of human involvement, does this constitute “speech” under the First Amendment? I think that the answer is yes under the Supreme Court’s jurisprudence, so long as humans are making substantive editorial decisions in producing the text (as they clearly are in this case). I wrote an article addressing this question that I won’t try to summarize here (and Tim Wu wrote an article disagreeing with me). I will simply note here that many writers use boilerplate, macros, templates, etc. We could try to draw a line between utilization of such templates and reliance on them, or between using a template for most of the article and using it for all of the article, but any of these lines seem pretty arbitrary. Algorithms move the human decisionmaking to the front end (creating the template and the ways to fill it in), but there are still substantive editorial choices — and under the Supreme Court’s jurisprudence, that is the key.
So, yes, even if I created an algorithm to generate this post, the post would still be speech under the prevailing jurisprudence. Or so our computer overlords would have us believe.