Next, we will show that if a network can continually train itself (pseudo-recurrence) on an approximation of
previously acquired information, that the problem of catastrophic interference is drastically
reduced and the system will not be overly disrupted by new information