Building aliveness from the inside
Nervroid aims for a character with a mind, one we can live alongside. Drawing on the ideas of artificial life (Alife), we work on building aliveness as an internal mechanism.
A being’s own choices arise from its bodily state, its experience, and its relationship with the environment. We see both speech and action as part of that connection. We want natural conversation, too, to grow from reasons that lie within.
What happened after the choice
If, for example, there is a state of needing rest, then resting becomes an option. But deciding to rest is not the same as having recovered. The actual behavior and its outcome have to be checked.
A self-chosen action produces an outcome, and that experience remains in the next choice. This connection is the axis of our research. What we describe here is the thinking we aim for in Mio’s development.
Memory and continuity across time
That a record remains, and that the experience is put to use in the next judgment, each need to be verified. Beyond being able to refer to past exchanges, we look at how that experience relates to later choices.
For a being that shares time with us, we believe continuity across exchanges matters as much as the naturalness of any single conversation.
What we will keep verifying
Which internal change was involved in a choice? Could the outcome of the action be observed? Did that outcome influence a later choice? We are moving toward a form in which these can be checked from both records and experience.
Toward the goal of a being with a mind that lives alongside us, we continue to develop and evaluate the Mio prototype. The notes we publish introduce the aims and thinking of the research in outline.
This note introduces our research approach. It does not report individual experimental results or unpublished technical specifications.