Sean Welsh

Sean Welsh

PhD Student
Department:

Sean is a PhD student from Australia, doing research in the area of robot ethics. As an undergraduate he majored in Philosophy at the University of NSW. He then embarked on a varied career including stints as an English Teacher, a tourism operator, an online travel agent, a software developer and a media adviser to a Federal politician in Australia before starting his PhD.

The working title of his doctoral dissertation is Moral Code: Programming the Ethical Robot. The central question of his research is: how could a robot make a moral decision?

His answer to this question (so far) involves three major elements.

First, there needs to be an ability for the robot to sense and classify data that is morally relevant. Technologies such as pattern recognition, object recognition and event recognition can be used to collect this data. The term lexical circuit can be used as an umbrella term to cover whatever sensory and cognitive functionality is needed to establish a well-formed formula (wff) of predicate logic e.g. Speeding(car). The end result of such “lexical circuits” is wff input into the “ethical governor,” the module of robot cognition that makes moral decisions. For example, Speeding(car) implies a moral action, namely, the issuing a speeding ticket to the owner of the car.

Second, there needs to be an ability to represent moral knowledge. Moral knowledge representation can be implemented in the form of axioms of deontic logic, a dialect of predicate logic with some “deontic” additions.

Third, there needs to be an ability to engage in moral reasoning, especially for more complex moral problems that involve clashes of duty and competing values. Simple moral problems can be solved with relatively straightforward automated theorem proving and simple lexical circuits. More complex moral problems will require more elaborate solutions involving graph databases, description logic and “semantic web” technology as well as axioms of deontic logic and lexical circuits.

Sean’s research sits on the intersection of the humanities and technology. It involves numerous controversial and interesting areas such as the ethics of military robots, the “correct ethical theory,” the claims that there can be “moral knowledge” and that such can be represented in the cognition of a robot and, finally, the nature of moral responsibility and whether robots could ever have it.