Reading Professor Jerry Kaplan’s “AI’s PR Problem” this morning made me think back to a talk I gave on a similar topic at Stanford Law School in 2010.
Let’s look back. (In what now seems like eons ago) I described a proto-taxonomy, one that distinguishes between AI entities that perform tasks that are inherently (i.e., normatively) the domain of natural persons and those that can be comfortably performed by non-humans. The former I labeled as Human Centric Tasks (HCT) and the latter as Entity Neutral Tasks (ENT).
Professor Lawrence Solum’s legal trustee application was the example I offered for an HCT. Apple’s Siri (though back then it was still a CALO product) was the example I provided for an ENT. This taxonomy also accommodates non-human AI derivatives, the “suprapsychological AI” mode of analysis (identified by Professor Owen Flanagan).
The HCT/ENT taxonomy helps dispose of the intransigent, obsessive pop-culture bred narrative of dangerous cyber killers. It helps fix AI’s PR problem. It also helps build a relevant legal framework that will have to deal with the AI ecosystem.