The Perfect Soldier

Email Print


The U.S. government
is looking for "An army of one," and for soldiers who
will "Be all that they can be." Unfortunately for the
Pentagon, all soldiers have the same defect: they are human. Humans
have emotions, fears, morals and worst of all — a family who will
ask questions when they die.

The Pentagon
has a terrible solution for this human problem — and it's straight
out of a science fiction movie. By
2010, they will invest $4 Billion into researching a fully automated
robot soldier.
According to scientists working on the project,
the goal is to program these machines to "not violate the Geneva
conventions" and “perform more ethically than human soldiers.."

The truth is
certainly more nefarious. There are few limits to the lengths that
the government will go to in enforcing U.S. policies. The state
engages in torture, illegal wars, and "black ops." These
policies reflect the wishes of our rulers, who are only restricted
by the capacity of humans to follow orders and keep quiet. It should
come as no surprise that the state would seek to remove the human

The Pentagon
tells us that their robots will make decisions which are based on
hard rules — not emotion, fear or vengeance. They say these robots
will perform more ethically than humans. To believe that this is
the Pentagon's true motivation, we'd have to ignore history and
common sense. Both tell us that we'd be naïve to take the government
at face value.

Let's assume
for a moment that the Pentagon's statements are truthful. It is
foolish to believe that these robots would be defect free. Remember
— these defects would exist in a system designed to take human life.
There are many risks with this plan which should be considered.

Creating a
robot soldier would be an extremely complex problem in software
engineering, wrought with the chance for software defects and logical
errors. For instance, consider the behavior this machine would need
to replicate. Humans take for granted most actions they perform
on a daily basis — however, these actions require complex mathematical
calculations and algorithms to replicate on a machine.

Think about
all the steps necessary to use a simple handgun. Assuming the weapon
is holstered and already loaded: you would remove the gun from the
holster; bring your arm up to point the weapon at the target — then
aim and fire.

For a robot
to replicate these actions would require thousands of instructions.
Consider all of the calculations not described which are easily
performed by humans — including identifying and classifying the
target, and the calculations necessary for aiming with some accuracy
(which must take into account weather conditions, terrain, wind
speed, the movement speed of the target, distance to the target,
etc.). Replicating this with software is not a trivial exercise.

It would be
foolish to think the software which controls this robot soldier
could be defect free. Writing code is a complex task — and humans
are not perfect. There is no amount of testing that would expose
all of these defects prior to the first field use of a given system.
If you've used any software product, you know this: whether it be
Windows, Office, or an ATM machine.

We can find
examples of defects being exposed in the field by looking at existing
military projects — such as the Patriot Missile system. The Patriot
system was designed to "provide a coordinated, secure, integrated,
mobile air defense system." Of course, this is a complicated
way of saying that the goal is to shoot down enemy missiles.

However, the
Patriot system was far from bug free. These problems were hidden
during military operations — state propaganda lauded the success
of the Patriot during the Gulf War. Only when military operations
ended did the true story become public.

President Bush claimed that the missile was responsible for the
downing of 41 out of 42 Scud missiles. Experts later refuted this
highly publicized success rate. In response to testimony and other
evidence presented, the staff of the House Government Operations
Subcommittee on Legislation and National Security reported, “The
Patriot missile system was not the spectacular success in the Persian
Gulf War that the American public was led to believe. There is little
evidence to prove that the Patriot hit more than a few Scud missiles
launched by Iraq during the Gulf War, and there are some doubts
about even these engagements. The public and Congress were misled
by statements of success issued by administration and Raytheon representatives
during and after the war."

The Patriot
failures in the Gulf war include a tragic loss of life on February
25, 1991. The Patriot system launched a missile intended to take
down a SCUD fired into Saudi Arabia. Due to a mathematical error
in the software, the Patriot missile was fired off target and missed.
The SCUD missile did not miss its target, resulting in the death
of 28 soldiers at the U.S. Army base in Dhahran. The error was a
miscalculation in the time that had elapsed since the system was
booted. As
noted here:

the time in tenths of seconds as measured by the system’s internal
clock was multiplied by 1/10 to produce the time in seconds.
This calculation was performed using a 24 bit fixed point register.
In particular, the value 1/10, which has a non-terminating binary
expansion, was chopped at 24 bits after the radix point. The small
chopping error, when multiplied by the large number giving the
time in tenths of a second, lead to a significant error. Indeed,
the Patriot battery had been up around 100 hours, and an easy
calculation shows that the resulting time error due to the magnified
chopping error was about 0.34 seconds. (The number 1/10 equals
In other words, the binary expansion of 1/10 is 0.0001100110011001100110011001100….
Now the 24 bit register in the Patriot stored instead 0.00011001100110011001100
introducing an error of 0.0000000000000000000000011001100… binary,
or about 0.000000095 decimal. Multiplying by the number of tenths
of a second in 100 hours gives 0.000000095100606010=0.34.)
A Scud travels at about 1,676 meters per second, and so travels
more than half a kilometer in this time. This was far enough that
the incoming Scud was outside the “range gate” that the Patriot
tracked. Ironically, the fact that the bad time calculation had
been improved in some parts of the code, but not all, contributed
to the problem, since it meant that the inaccuracies did not cancel."

This is far
from the only problem with this system. Here are further examples:

Remember that
these errors occurred with a "defensive" system. The Pentagon's
robot soldier will not be for defensive purposes only. This machine
will actively decide whether or not to end a human life. It will
involve more complex operations than the Patriot — operations which
will require massive research and development of an artificial intelligence.
We have to assume that problems with a robot soldier would dwarf
those encountered with the Patriot missile system.

The Pentagon
claims these robots will be designed to conform to Geneva conventions.
Still, who will pay the price if a logical software error causes
the robots to commit a war crime — such as misidentifying a village
full of children as a target which must be eliminated. Will the
government behave like Microsoft and issue a "hot fix"
for this bug?

The problems
with this plan are not just technical. We should not trust the government's
motives for pursing this project — there are many obvious benefits
they have not mentioned.

One immediate
benefit of this robot is that it will remove a restraint on the
government's ability to conduct war at a large scale — that is,
the public criticism surrounding the loss of American lives. This
benefit will certainly cause the project to gain a great deal of
favor among Pentagon officials.

we'd likely see a minor rollout of these robotic troops for "high
risk" situations. Eventually the government would systematically
replace all front line troops. Only human support staff would be

Consider the
occupation of Iraq — dangerous zones could be patrolled by these
robots. They would certainly be more durable than humans, requiring
no food, shelter or sleep. Only a power source to recharge their
batteries and access to the central system would truly be necessary.

The public
would have a completely different view of the war with the risk
to American soldiers largely removed. The public and media generally
ignore the effects of war on those we occupy. The potential for
American military destruction abroad will not only increase, but
it will do so at the hands of a soldier which will not confess details
later on.

There are other
benefits as well. Today, military leadership must concern itself
with the size of the force, the negative PR associated with operations
as well as certain programs (i.e. "stop loss"). These
concerns will decrease or cease to exist when humans are taken out
of the equation.

Once the Pentagon
is able to mass produce these metallic beasts, there will be few
restraints on their actions. In the long run, it will become easier
to engage in war.

The logical
next question is: "Why would the government only deploy this
technology to the military?"

I suspect it
wouldn't be long before these troops patrol the streets of America
as well. The public may be told that dangerous inner city neighborhoods
can be cleaned up without risk to the police force. Robots could
replace SWAT teams, and even regular police officers.

This technology
would be able to take advantage of the state invasion of our privacy.
Law enforcement robots will have immediate access to all of your
records, recent transactions, and other private data through existing
government and civic systems. You don't have to be paranoid to see
the potential for further government intervention in your life.
These tracking mechanisms are already being put in place — think
of RFID tags and Real ID. Today, it may sound like science fiction:
a robot police force could arrest or fine a citizen based on a set
of weighted criteria — using data gathered from these systems.

The total scenario
described here may not to play out in the short term. There is still
a great deal of research necessary. However, the Pentagon has committed
a huge sum of money towards the goal of an autonomous soldier, and
they've deployed remote control robots which are human controlled
(i.e. the Predator drone). It is closer than you think.

This advancement
will be celebrated at home for the U.S. troops it will save. Technological
advancements have enriched our lives in many ways. However, the
potential for abuse at the hands of our government is too great
to ignore. Instead of focusing on the "Terminator" scenario
(robots turning against the human race), we should focus on the
real threats posed by this technology. Are we willing to accept
a machine which determines whether or not to end a human life?

22, 2008

Sica [send him mail]
is a software engineer living in Woodbury, Connecticut, where he
volunteers as a precinct leader with the Campaign
for Liberty

Email Print