Agents and Devices: A Relative Definition of Agency
This work provides a theoretical framework for understanding agency in AI, but it is incremental as it builds on existing philosophical concepts like Dennett's stances.
The paper tackles the problem of formally distinguishing between devices and agents in computational systems by proposing a definition based on input-output mappings versus optimization functions, and demonstrates this using Bayes' rule on behavior in a toy grid-world.
According to Dennett, the same system may be described using a `physical' (mechanical) explanatory stance, or using an `intentional' (belief- and goal-based) explanatory stance. Humans tend to find the physical stance more helpful for certain systems, such as planets orbiting a star, and the intentional stance for others, such as living animals. We define a formal counterpart of physical and intentional stances within computational theory: a description of a system as either a device, or an agent, with the key difference being that `devices' are directly described in terms of an input-output mapping, while `agents' are described in terms of the function they optimise. Bayes' rule can then be applied to calculate the subjective probability of a system being a device or an agent, based only on its behaviour. We illustrate this using the trajectories of an object in a toy grid-world domain.