3, the First Doctor and his companion, Dodo Chapman, return to 1960’s London in
the episode The War Machines. After hanging an “Out of Order” sign on his
TARDIS so no one barges in a perfectly normal seeming police box, the Doctor
notes there’s something strange going on.
They track the Doctor’s strange feeling back to the Post Office Tower,
where he and Dodo stumble upon WOTAN (Will Operating Thought ANalogue). WOTAN is one of the first artificial
intelligences on Earth, and it has decided to use that artificial intelligence
to hypnotize humans, create an army of robotic warriors and use this army to
take over the Earth. He and his robotic
warriors called War Machines decide they are more deserving of Earth as humans
are inferior creatures. Dodo and the
Doctor, as well as his new companions Polly Wright and Ben Jackson, eventually
stop WOTAN. To stop WOTAN the Doctor had
to capture a War Machine and reprogram it to destroy WOTAN. But what if WOTAN could have been stopped and
spared destruction. Could the Doctor
have inflicted meaningful punishment on an artificial intelligence and its
punishment argue that punishment should be done out of one of three possible
motives: retribution, rehabilitation
or deterrence. If the Doctor wanted to punish WOTAN for his
wrongdoing as a form of reparation for WOTAN’s wonton violence and return the
Universe to some sort of cosmic balance, he would have been acting out of a
retribution motive. A rehabilitation
defense of punishing WOTAN would be using punishment designed to change WOTAN’s
thinking so it no longer wants to participate in his plans for worldwide
domination. An attempt to create
deterrence through punishing WOTAN would be to persuade WOTAN (and his evil
supercomputer compatriots) from engaging in hypnotism and creating a robot army
of domination in the future.
these ideas of punishment require that the punished have agency over their actions.
Agency means that an agent has control over their actions because (1) their
actions stem from their inner nature (2) they are not under the control of
something external to themselves. When
Dodo conspires to have the Doctor kidnapped, she is not acting as an
agent. Her actions are not consistent
with her inner nature of respect for the doctor because she is under the hypnotic
control of WOTAN. The War Machines
probably aren’t agents when they go on a rampage. While they may be acting in a way consistent
with their inner nature (or rather programming), they can’t be said to be not
under the control of something external to themselves. Their programming was established external to
themselves and they are merely following its orders. But WOTAN can be accurately described as an
agent. It seems to believe that humans
truly are inferior to machine, and so its actions seem to be consistent with
this. Furthermore, it isn’t under the control
of another, it makes its own decisions.
One could argue that because it is programmed (similar to the War
Machines and many other computers and robots) and so under the control of
another, it isn’t an agent. But it’s clear
during the War Machines serial that
WOTAN has surpassed its programming and truly decided to do what’s best for its
But each of
the justifications for punishment requires something besides simply
agency. In order for retribution to make
sense WOTAN has to also be able to suffer so that it can suffer in fair
proportion to the harm it has caused.
But can a supercomputer be made to suffer? If I remove one of the wheels on WOTAN’s
exterior that it uses to run some program, does WOTAN suffer harm as a result? Is it a mere annoyance to WOTAN that it needs
to find a workaround for? Or does it
even notice and care that it’s lost the wheel?
If these questions can’t be answered in the positive, then it probably isn’t
the case that WOTAN suffered and retributive justice doesn’t make sense so we
shouldn’t even bother.
and deterrence both require an ability on WOTAN’s part to be able to
learn. If the agent can’t learn then it
can’t be learn such actions are bad and incur punishment. If one of the other computers in Post Office
Tower crashes, hitting it to teach it a lesson not to crash again won’t do any
good. It can’t learn that it’s bad to crash because it doesn’t have that
capability. WOTAN, as an example of artificial
intelligence, has the prerequisite ability to learn. It can learn that if it is punished (however
this punishment is made to happen), the actions are bad and should not be done
again. If it witnesses a similar supercomputer
that tries to take over the world and fails before being punished, it may learn
not to even bother attempting its global domination. Of course it may be more beneficial to
program artificial intelligences to learn and correct their behavior without punishment.
the Doctor choses the ultimate punishment for WOTAN – the “death” penalty. We’ve previously looked at if we can justify
the death penalty for Blon from Raxacoricofallapitorious, but can such arguments
be made to justify the Doctor destroying WOTAN?
It’s difficult to argue that the Doctor was trying to make a
retributivist justification of stopping WOTAN.
It’s probably more likely that by destroying WOTAN the Doctor was acting
on a rather literal interpretation of deterrence theory. If WOTAN is destroyed then it can no longer
command the War Machines and kill people.
It has been deterred from further violence in that it is incapable of
further violence. This may be the best
way to justify punishing robots. A broken
machine, even a megalomaniacal, genocidal, world-domination seeking one, can’t
commit crimes if it doesn’t exist anymore.
This approach certainly one that the Doctor prefers to take in his
dealings with murderous machines in his future adventures. He doesn’t seem to overly concern himself
with rehabilitation and offering machines second chances. He knows that machines can feel and learn (given
his relationship with his TARDIS) but that doesn’t seem to change much for him.