Figure 2.

Legend: A focus on the size or scale of goals any given system can pursue, as an invariant across the space of possible sentient beings of whatever embodiment, allows plotting very diverse intelligences on the same graph [23]. The scale of their goal-directed activity is estimated (collapsed onto one axis of space and one of time, as in space-time diagrams). Importantly, this way of visualizing the sophistication of agency is a schematic of goal space—it is not meant to represent the spatial extent of sensing or effector range, but rather the scale of events about which they care and the boundary of states that they can possibly represent or work to change. This defines a kind of cognitive light cone (a boundary to any agent’s area of concern); the largest area represents the “now”, with fading efficacy both backward (accessing past events with decreasing reliability) and forward (limited prediction accuracy for future events). The diamond or “spinning top” shapes of the cones depicted above are simplifications. Agents are compound entities, composed of (and comprising) other sub- or super-agents, each of which has its own cognitive boundary of various sizes. Image by Jeremy Guay of Peregrine Creative. Selves increase their cognitive boundary by connecting together (“GJ”, standing for gap junctions—an example of a biophysical connection used by cells to merge into higher-level beings) in functional ways that allow simple homeostatic loops to measure, implement, and remember progressively larger states (thus increasing the scale and complexity of what they Care about). Top panel used by permission, produced by Jeremy Guay of Peregrine Creative.