Error-Correction for AI Safety
Leon Kester2020, Artificial General Intelligence
Sign up for access to the world's latest research
Abstract
The complex socio-technological debate underlying safetycritical and ethically relevant issues pertaining to AI development and deployment extends across heterogeneous research subfields and involves in part conflicting positions. In this context, it seems expedient to generate a minimalistic joint transdisciplinary basis disambiguating the references to specific subtypes of AI properties and risks for an error-correction in the transmission of ideas. In this paper, we introduce a high-level transdisciplinary system clustering of ethical distinction between antithetical clusters of Type I and Type II systems which extends a cybersecurityoriented AI safety taxonomy with considerations from psychology. Moreover, we review relevant Type I AI risks, reflect upon possible epistemological origins of hypothetical Type II AI from a cognitive sciences perspective and discuss the related human moral perception. Strikingly, our nuanced transdisciplinary analysis yields the figurative formulation of the so-called AI safety paradox identifying AI control and value alignment as conjugate requirements in AI safety. Against this backdrop, we craft versatile multidisciplinary recommendations with ethical dimensions tailored to Type II AI safety. Overall, we suggest proactive and importantly corrective instead of prohibitive methods as common basis for both Type I and Type II AI safety.
Figures (2)
![Fig. 1. Taxonomy of pathways to dangerous AI. Adapted from [37]. references to assumed AI risks explicit. For this purpose, we employ and subse- quently extend a cybersecurity-oriented risk taxonomy introduced by Yampol- skiy [37] dis modifying it clustering o played in Figure 1. Taking this taxonomy as point of departure and while considering insights from psychology, an ethically relevant f systems into Type I and Type IT systems with a disparate set of properties and risk instantiations becomes explicitly expressible. Concerning the set of Type representing the complement of the set of Type II systems. Conversely, Type II systems as systems with a scientifically plausible ability to hypothetica systems of which present-day Als represent a subset, we d efine it as we regard act independently, intentionally, deliberately and consciously. Given the contro- versial ambiguities linked to these attributes, we clarify our idiosyncratic use with a working definition for which we do not claim any higher sui general, but Type II systems, which is particularly conceptualized for our line of argum tability in ent. With we refer to systems having the ability to construct counterfac- tual hypotheses about what could happen, what could have happened and why including the abi cluding the possibility of what-if deliberations with counterfactual de ity to simulate “what I could do”, “what I could have done” and the generation of “what if’ questions. (Given this conjunction of a bilities in- pth about self and other, we assume that Type II systems would not represent philosophi- cal zombies. A detailed account of this type of view is provided by Friston in [20] stating e.g. that “the key difference between a conscious and non-conscious me is that the non-conscious me would not be able to formulate a “hard problem”; quite simply because I could not entertain a thought experiment” .)](https://figures.academia-assets.com/89101063/figure_001.jpg)
![Fig. 2. Transdisciplinary system clustering of ethical distinction with specified safety and security risks. Internal causes assignments require scientific plausibility (see text). to the risks [7g and Ih for the former as well as [Jt and IT) for the latter subcat- egory respectively. The reason for augmenting the granularity of the taxonomy is that since Type II systems would be capable of intentionality, it is consequent to distinguish between internal causes of risks resulting from intentional actions of the system and risks stemming from its unintentional mistakes as parallel to the consideration of external human-caused risks a and b versus c and d in the matrix. (From the angle of moral psychology, failing to preemptively consider this subtle further distinction could reinforce human biases in the moral percep- tion of Type II AI due to a fundamental reluctance to assign experience [25], allibility and vulnerability to artificial systems which we briefly touch upon in Section 3.2.) Especially, given this modification, the risks [7g and Ih are not necessarily congruent with the original indices g and h, since our working def- inition was not a prerequisite for the attribute “independently” in the original taxonomy. The resulting system clustering is illustrated in Figure 2.](https://figures.academia-assets.com/89101063/figure_002.jpg)
