A matter of learning paradigm
Both Terminator and Robocop are very high tech systems, representing
a big challenge for the future research on advanced autonomous robots.
The main distinguishing feature between them is that Terminator is a
pure artificial robot, whereas Robocop is a hybrid (biological-robotic)
system: a mechatronical body controlled by a modified human brain, in
which most of the low-level functions have been replaced by dedicated
processor chips.
Another big difference between these two robots concerns the type of
learning paradigm adopted in their control processes. For example,
Robocop's tracking controller uses a supervised learning paradigm,
whereas the Terminator's neural net processor can work under different
learning schemes.
Robocop's learning paradigm
In one scene of the Robocop movie, we see that, after a long
shoot-out, Robocop reports some damage in the eye-arm calibration
system, loosing the capability of tracking targets with his gun. For
this reason, he goes in a derelict building and asks his assistant to
take aim with his arm and shoot a target. In other words, Robocop shows
that he is not capable of self-learning, but he needs a supervisor (a
teacher) that shows him the correct examples.
Terminator's learning paradigm
On the other hand, Terminator learns from his own experience.
We see that he learns John Connors's language expressions and
behaviors. For example, Terminator learns from John to flip down
the sun visor of the car to find the keys for starting the engine
(before, he broke the cowl around the steering column to join the
wires).
