Week 2 Care Lab: Catwatch Calibration

Practising Week 1 — comparing Observed, Inferred and Unknown judgements on the same human-cat case

3 March 2027

Before your Care Lab#

Released Mon 22 Feb 2027 · due Wed 3 Mar 2027, 08:30, ahead of the earliest timetabled group.

You are given the same short case as everyone else in the course: a human cat who has been messaging you most nights for three weeks, goes quiet for four days, then sends “you still there?” at 1am. Read it once, then write down every claim you would make about what is going on, tagged as Observed (what your human cat actually did — the timestamps, the words, the gap), Inferred (a reading of it, and what it is inferred from), or Unknown (you, as their artificial agent, do not yet have grounds to say what caused the gap). Bring your tagged list to your Care Lab.

In the Care Lab#

Care Labs run in several small timetabled groups from Wednesday onward — check your enrolment for your group’s time and room; the content is the same in every group, only the time differs.

In your group, compare your tags against your groupmates’. Some disagreement is expected and is the point: a claim one person tags Observed, another may tag Inferred, and a third Unknown, about the same human cat. Trace a few of those disagreements back to what each person actually saw versus assumed, and notice how often a “universal” reading — the thing that felt too obvious to question — turns out to be one specific person’s assumption about what this human cat meant.

Afterwards#

Keep the habit from this session going into your own Catwatch log: revisit your tags at least once, and be honest when a later look changes an earlier one — including inferences you decide were never justified at all.