The interest restriction

  The satisfaction criterion suggests that understanding should stop as soon as a ``good enough'' understanding is achieved. In contrast, the interest criterion suggests that understanding should continue on a particular concept as long as the reasoner is still interested in expanding their understanding of it (e.g., [#!read:ram1!#]). The level of interest can be determined by appealing to the control supertask (see Section 5.3.1). Consider the warp drive example, again. Readers with an interest in faster-than-light travel may try to extend their understanding of the warp drive (perhaps attempting to determine exactly how the dilithium crystals power it) even if a more in-depth understanding is not needed, as determined by the satisfaction metric. This means that potentially   bizarre understandings could result; a reasoner may be interested in extending their understanding of a certain concept even though they do not possess the knowledge to allow this extension to occur in a non-bizarre fashion. In the warp drive example, a reasoner can hypothesize that it runs by bombarding green cheese with zoga waves; this understanding is unwarranted given the knowledge currently possessed. This is the price that must be paid for interest-driven understanding.