Download Today, the usage of artificial intelligence in military

Running Head: APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
Abdul Muhaimin Abdul Rao
Khairi Mirza Khairuddin
Muhammad Ariff Abd Rashid
Running Head: APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
Table of Contents
1.0 Introduction .......................................................................................................................... 1
2.0 Advantages ........................................................................................................................... 3
3.0 Disadvantages ...................................................................................................................... 5
4.0 Conclusion ........................................................................................................................... 7
5.0 References ............................................................................................................................ 8
1
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
1.0 Introduction
The development of technologies is now being concerned by people about what sort of
world they would live in the next 20 years ahead. Maybe a world that full with robots like most
of the science fiction movies such as Terminator that shows humans and robots are living
together in the future. By looking at the advancement of technologies now, it is not impossible
that science fictions will become reality. It all started since 1950s where there was a rapid
growth of mathematical logic and new ideas about computation which then leads to the
evolution of Artificial Intelligence (AI). Mathematical logic became a site for researchers to
study specifically about AI (Barr & Feigenbaum, 1981). (McCarthy, 2007)suggested that
people began to work independently on intelligent technologies after World War II and Alan
Turing, an English mathematician may have been the first person to work on this and decided
that AI was best researched through programming computers rather than by building machines.
By the late 1950s, many AI researchers were taking programming computers as a base for their
works.
First of all, artificial intelligence is known as a computer program that is operating by
its own will without being controlled by humans or also known as autonomous. But basically
they are originally created by humans themselves and have been programmed to help people
with their works. Copeland (2000) defined AI as an area of science creating computing devices
to work using understanding and intelligence when otherwise actioned by humans. Another
definition of AI is smart devices such as intelligent computer programs are linked to the same
task of using computers to analyse intellectuality of people. However, AI does not need to
restrict itself to methods that can be observed physically (McCarthy, 2007).
Today, the usage of artificial intelligence in military is quite significant in their training
and also in a real war. Conant (2013) said autonomous systems have begun to give an important
effect for military in warfare. Unmanned drones are used to provide sustained surveillance,
accurately attack on high-value targets and small robots are used for tasks to deactivate
explosive devices. These systems are generally piloted by humans and rely on near-continuous
control. Other than that, there will be a great test for AI researchers in 2016 to find out how to
make a machine to think about itself or in other words to make it autonomous. The military
hopes that AI will be able to control everything from cyber soldiers to the target systems in the
jet fighters, F-35 (Bose, 2015).
2
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
Ultimately, capability of self-aware and being adaptive are some of the criteria that
should be in an intelligent machine. Crowder (2016) said a real AI machine should think and
live like a living organism which can develop and adapt with its surrounding. Robotic
cockroach created by John Carbone and James Crowder himself is an example of intelligent
device that can live autonomously, adapt itself to various conditions and aware of its
surrounding. It is created to have a distributed intelligence system similar to the distributed
brain system of an octopus.
The same specification is being used by the military by setting up a goal to develop a
system that is able to perform a variety of autonomous behaviours under different situations of
battlefield (Conant, 2013). For an example, according to Davenport (2016) in his article,
Prabhakar said Defence Advanced Research Projects Agency (DARPA)’s 132-foot-long
autonomous ship is one of the example of AI machines that worked without any crew. It can
pass in hostile territory without having possibility of casualties and it is also designed to travel
thousands of nautical miles.
However, capabilities of artificial intelligence in military are questioned by some
parties by claiming that the application of AI in military is not advisable in the future. This is
because the lack of emotional of an artificial intelligence machine will lead to the difficulty in
understanding humans desire. For another reason, McCarthy (2007) explained the ethical
dangers or drawbacks from trying to create AI devices. He questioned what if the smart devices
become more intelligent than humans and are they willing to work for humans? The biggest
threat is when the intelligent machines become killer robots which would kill thousands of
people.
Last but not least, by looking at the discussion of advantages and disadvantages of
artificial intelligence in military later, it is accepted that in the future AI should not be applied
for military usage. Comparison between pros and cons of AI will bring to a conclusion for the
applicability of artificial intelligence in military in the future.
3
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
2.0 Advantages
Most of autonomous robots - robot that performs tasks with a high degree of autonomy
have the ability to do job that cannot be done by human beings or even impossible to be done
by any of the creature that surround this world. Being in the extremely hot or cold condition,
searching and rescuing victims at radiation site and exploring undersea which is more than 400
feet deep are examples of impossible things to be done by normal human being. In these cases,
autonomous robots are needed to overcome human incapability to be at those extreme places
because it is too peril to be at those places under that condition without any protective
equipment. Even if human being can withstand under those condition, the cost of those
equipment would be higher than the cost to build a robot which is capable to do the same task
(The Engineer, 2015).
However, there are still some things to be taken into consideration when they are
dealing with those environments. For an instance, any substance even a metal may melt if the
surrounding temperature is higher than boiling point of the metal itself. To solve this problem,
a metal which have a high melting point should be considered in the making of the body of the
robot. In term of defence sector, Shaun Whitehead who is the founder and director of Scoutek
said that some robots are available for extreme conditions but the biggest deal is not the whole
part of the robot’s body but rather important individual component lies within the body of the
robot itself. Radiation hardening is one of the efforts in protecting the important components
from being damaged which is in this method the electronic components is protected from any
ionizing radiation which may cause malfunction.
In military, it is crucial and essential to survive in such extreme condition whether
during a warfare or search and rescue. Therefore, autonomous robots intended to be used in the
battlefield as to overcome human limitations and to enhance the capability of the army. To
begin with, robots have already served human for decades in many situations. For example,
robot autopilot called Elmer which is named after an American inventor flew bombers during
World War II and an unmanned aircraft had been flying at least since Vietnam War (Lloyd,
2015). In addition, seafloor was being mapped routinely for at least twenty years by
autonomous underwater vehicles which human would not capable to withstand high pressure
(MBARI). As a matter of fact, recently, Navy have opened the facility that features a wide test
areas designed for autonomous robots to operate in the real world condition such as in coastal
4
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
waters, deserts and jungles which is located at the Naval Research Laboratory in Washington,
D.C (Weinberger, 2012). This development would give a positive outcome of applicability of
Artificial Intelligence in military.
There are a lot of differences between human and robot and the most obvious part is
the presence of failures. Compared to autonomous robots, humans have mental limitations that
lie within them that can only be explained through biology. In battlefields to be exact, there are
some advantages of using machines instead of human army. Machines such as autonomous
robots do not feel exhausted easily, do not blink, do not look for protection under the trees
during rainfall and do not chitchat with the other robots (BBC, 2002). All of these are flaws
that human beings have but robots do not. In war, these flaws must be overcome by any means
so that the opponent would not have the advantage to win over the battle. Since robots do not
own such flaws, the chances of winning the war may be brighter.
As for the first reason on why machines are preferred more rather than human in
battlefield is because machines do not feel fatigue easily. On the other hand, the duration taken
for a normal human being to get their job done without feeling fatigue fully depends on his
stamina. Stamina is inconsistent and dependent on the lifestyle of the individuals. This
inconsistency must be avoided in military but the possibility for this to happen is almost
impossible. Exercising regularly may help the army to be at the top of their performance
without exhaustion (Talent, n.d.). The military saying physical fitness is the cornerstone of
combat readiness (Talent, n.d.) proves that exhaustion should be avoided in military by
replacing human army with military robots as robots cannot feel any exhaustion and can do
lots of things without breaking a sweat
Secondly, there is not much restriction on what situation that a robot can work at its
best as it can work in almost every weather based on what it is designed for. In robotics
designing process, the designer should gather information regarding the environment in which
the robot is to be placed in and consider several matters on how the robot would get their job
done such as the movement of the robots and the equipment needed by the robots (Robotics,
n.d.). However, in military to be specific, an army needs to cope with the bad weather rather
than the harsh environment of the battlefield itself. For human armies, looking for a shelter
during those weathers is necessary so that the battalion will not fall sick. Oppositely, for
unmanned ground vehicle (UGV) which is one of the military robots, there is an autonomous
5
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
hunter-killer team that is designed to do a sneak attack on the enemy frontline despite of the
bad weather (Pejic, 2016).
Normal human being can only pay full attention to what he is doing for about eight
seconds in average (Static Brain, 2016), which means after the first 8 seconds of focusing onto
something, the human mind tends to wander off thinking about something that is less important.
In the military sector, each army must have the ability to stay focus at all times on the battle
field or else the enemy will get the better of them. Unfortunately, a typical human army is
unable to stay focus for too long as the human mind can be easily distracted either through
sensory distraction or emotional distraction (Cooper, 2014). The trend of human attention to
stay focused on guard drops dramatically at the first 30 minutes (BBC, 2002) but not for
military robot which will not have any drop of focus on what they are doing.
3.0 Disadvantages
The development of AI however has made fellow researches and public figures such as
Elon Musk, Stephen Hawking, and Nick Bostrom to show their concern towards the negative
impacts that AI could bring to humanity. Nick Bostrom who is a philosopher at the University
of Oxford has shown his worry regarding this issue through his book entitled
“Superintelligence: Paths, Strategies, Dangers”.
Via his book, he expressed his anxiety toward the humanity fate if human race
depending on Artificial Intelligent (AI) rather than our own ability to solve problems. He also
pictured the situation in his book on what would happen if AI is too powerful and we are too
depending on them. The fate of humanity would be the same like the fate of the gorillas
nowadays as the fact as the gorilla species nowadays is depending more on human rather than
the gorillas themselves (Bostrom, 2014).
This philosopher who is known for his work on super intelligence risk has compared
AI to nukes repeatedly in his book. Some people might be wondering why he has to compare
AI to nukes and not to anything else. Firstly, we all know how dangerous nukes are through
the atomic bombings on Hiroshima and Nagasaki on August 6, 1945 as it had caused a lot of
deaths of about 225,000 in total. Secondly, it is more applicable and understood to use nukes
as a comparison rather than comparing it to the AI itself which is quite complicated (Green).
6
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
Elon Musk, CEO of Tesla which is an American automotive and energy storage
company is at the same side as Bostrom. He, through his account Twitter has shown his concern
towards this issue by urging the society to be extra careful with AI because it has the potential
to be more perilous than nukes. Recently, he had invested in an artificial intelligence company
called Vicarious in order to inspect the progress of artificial intelligence. (CNBC, 2014)
On the other hand, academics believe that autonomous weapon empowered by artificial
intelligence would be the threat of an arms race and they agree to signed a letter warning for
banning on these technologies (Mims, 2015). In 2009, researchers and academics expert
discussed several matters about the potential impact of robots and computers and the effect of
hypothetical prospect that these man-made machines could become self-reliant and have full
power to make decision. The use of robots with high extent of capability in making decision
for military purpose was questioned by some academics. They also discussed about the degree
of catastrophe might be posed by robots and level of autonomy could be obtained by computer.
Therefore, some have suggested to build so-called “Friendly AI” which mean that the effort to
make AI profoundly friendly and humane are in the same level with occurring development of
AI (Ryan, 2014).
A decade ago, the existence of this technology not getting a lot of attention as it just
emerging onto the battlefield. However, as this technology is developing, groups of expert
believe that it would become disaster toward humanity (Lin, 2015). Killer robots are questioned
its reliability to kill someone without human intervention and these killer robots also seem to
be unfavoured by the laws of war. Central principles underlying laws of war (LOW) requires
that the war should be ended quickly, not causing any unnecessary destruction and civilian
should be protected against any destruction and poverty (Bolgiano & Patterson, 2011).
. In the meantime, artificial intelligence still could not distinguish friend from foe
reliably. Professor Noel Sharkey said that the technology for distinguishing between combatant
with guns and civilian is at least 50 years away. But, as a matter of time, this technology may
be developed in the foreseeable future and thus could comply one of the LOW. Second, selfdefence of the robot should be on the same level as human being. However, Pentagon insisted
that their Unmanned Aerial Vehicle (UAV) which is drone has the right for self-defence, which
mean that the machine could tactically fire on potential threats. It may sound logic but this
would lead to killing someone without ethical concern because they use self-defence as the
reason to kill violently (Singer, 2009).
7
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
Even though robots could follow all of the rules of engagement (ROE)-rules for military
forces which define the conditions, degree and manner to use deadly forces (Bolgiano D. ,
2007) ,it is not guaranteed that the machines could easily figure the enemy at the first place. It
could also become complicated if the enemy could hide among civilian. Then, this would
expose civilian to danger. For example, it is permissible to shoot the enemy if it is hiding behind
a child? Or what if the enemy is just plotting an attack but did not execute it yet? Can we shoot
the target? It is irrational to think that robots to find them any easier (Singer, 2009).
4.0 Conclusion
As a conclusion, most researchers are still trying to figure out the consequences of using
AI in military. In addition, AI will keep on developing and it is expected to achieve the peak
of the development thus it is not impossible for the AI to become main core of the military.
This is due to the fact that robot can withstand to be in any extreme environment which is not
suitable for human. Furthermore, the involvement of AI in military can enhance the current
army capability as they do not feel fatigue easily, able to adapt in any weather and can pay full
attention onto what they are doing. After discussing the advantages and also disadvantages of
using AI for military purposes, it seems that a group of academics had few opinions, some of
them agree by ensuring that this system is working smoothly even under any unexpected
situation, and the other group of academics are trying to offset the advantages by signing an
agreement to not involve AI in military because of the risk to humanity by their involvement.
These academics scared of the fact that AI could disobey the laws of war and in the worst cases
it could be more dangerous than nukes. However by the year 2045, it is expected that machines
in general will be able to do if not all work that human can do, then a very notable portion of
the work will be enough (Muehlhauser, 2013). If this prediction is true, then the world should
be fully prepared for any possible circumstances so that this event could benefit humanity
rather than causing trouble (Muehlhauser, 2013). Based on our point of view, we think that
human are responsible for the peacekeeping of each country. As we all know that human should
be the only one creature that is responsible in this planet Earth. We should not let go this
responsible to the something that we created. Finally, it is believed that the idea of applying AI
in military should be opposed as it is agreed by fellow academics and researches the
enhancement of full AI could sign the end of human race (Lewis, 2014).
8
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
5.0 References
Barr, A., & Feigenbaum, E. (1981). The Handbook of Artificial Intelligence (Vol. I).
Carlifornia: William Kaufmann, Inc.
BBC. (12 April, 2002). Robot soldiers. Retrieved 3 September, 2016, from BBC:
http://news.bbc.co.uk/cbbcnews/hi/teachers/citizenship_11_14/subject_areas/scientifi
c_development/newsid_1923000/1923299.stm
Bolgiano, D. (2007). Combat Self-Defense: Saving America's Warriors From Risk-Averse
Commanders and Their Lawyers. Little White Wolf Books.
Bolgiano, D. G., & Patterson, J. M. (2011). Fighting Today's Wars: How America's Leaders
Have Failed Our Warriors. Stackpole Books.
Bose, N. (17 December, 2015). US military wants to use artificial intelligence to create cyber
soldiers. Retrieved from Times News Daily: http://timesnewsdaily.com/us-militarywants-to-use-artificial-intelligence-to-create-cyber-soldiers-6444
Bostrom, N. (2014). “Superintelligence : Paths, Strategies, Dangers”. Great Clarendon
Streeet,Oxford.
CNBC. (4 August, 2014). AI potentially 'more dangerous than nukes,' Musk warns. Retrieved
2 September, 2016, from CNBC: http://www.cnbc.com/2014/08/04/ai-potentiallymore-dangerous-than-nukes-musk-warns.html
Conant, J. (10 January, 2013). Army researchers develop robot intelligence to support soldiers.
Retrieved
from
U.S.Army:
https://www.army.mil/article/94076/Army_researchers_develop_robot_intelligence_t
o_support_Soldiers
Cooper, B. B. (7 February, 2014). The two brain systems that control our attention: the science
of
gaining
focus.
Retrieved
4
September,
2016,
from
Buffer
Social:
https://blog.bufferapp.com/the-science-of-focus-and-how-to-improve-your-attentionspan
Copeland, J. (May, 2000). What is artificial intelligence? Retrieved from AlanTurning.net:
http://www.alanturing.net/turing_archive/pages/reference%20articles/what%20is%20
ai.html
9
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
Crowder, J. (16 June, 2016). Neural networks: artificial intelligence and our future. Retrieved
from TC: https://techcrunch.com/2016/06/16/neural-networks-artificial-intelligenceand-our-future/
Davenport, C. (17 June, 2016). Robots, swarming drones and ‘iron man’: welcome to the new
arms
race.
Retrieved
from
The
Washington
Post:
https://www.washingtonpost.com/news/checkpoint/wp/2016/06/17/robots-swarmingdrones-and-iron-man-welcome-to-the-new-arms-race/
Green, D. (n.d.). Why is AI thought to be more dangerous than nukes yet we still can't reach
the complete human level? Retrieved 2 September, 2016, from Quora:
https://www.quora.com/Why-is-AI-thought-to-be-more-dangerous-than-nukes-yetwe-still-cant-reach-the-complete-human-level
Lewis, T. (4 December, 2014). Artificial Intelligence: Friendly or Frightening? Retrieved 4
September, 2016, from Live Science: http://www.livescience.com/49009-future-ofartificial-intelligence.html
Lin, P. (20 April, 2015). Do killer robots violate human rights? Retrieved from The Atlantic:
http://www.theatlantic.com/technology/archive/2015/04/do-killer-robots-violatehuman-rights/390033/
Lloyd, P. A. (2015). The use of drones during the Vietnam war. Retrieved from Peter Alan
Lloyd: http://peteralanlloyd.com/general-news/the-use-of-drones-during-the-vietnamwar/
MBARI. (n.d.). Seafloor mapping AUV. Retrieved from Monterey Bay Aquarium Research:
http://www.mbari.org/at-sea/vehicles/autonomous-underwater-vehicles/seafloormapping-auv/
McCarthy, J. (12 November, 2007). Basic questions. What Is Artificial Intelligence?, 2-4.
Mims, C. (17 August, 2015). We’re fighting killer robots the wrong way. Retrieved from The
Wall Street Journal: http://www.wsj.com/articles/were-fighting-killer-robots-thewrong-way-1439784149
Muehlhauser, L. (15 May, 2013). When will AI be created? Retrieved 4 September, 2016, from
MIRI: https://intelligence.org/2013/05/15/when-will-ai-be-created/
10
APPLICABILITY OF ARTIFICIAL INTELLIGENCE IN MILITARY
Pejic, I. (2 February, 2016). Military analysis: U.S. UAV’s, UGV’s, autonomous weapon
systems and military robotics. Retrieved 4 September, 2016, from SOUTH FRONT:
https://southfront.org/military-analysis-u-s-uavs-ugvs-autonomous-weapon-systemsand-military-robotics/
Robotics. (n.d.). Robotis design process. Retrieved 4 September, 2016, from Robotics:
http://www.galileo.org/robotics/design.html
Ryan, M. (2014). The Digital Mind: An Exploration of artificial intelligence.
Singer, P. W. (2009). Wired for War: The Robotics Revolution and Conflict in the 21st Century.
Penguin.
Static Brain. (2 July, 2016). Attention span statistics. Retrieved 4 September, 2016, from Static
Brain: http://www.statisticbrain.com/attention-span-statistics/
Talent, E. (n.d.). Importance of exercise in the military. Retrieved 3 September, 2016, from
acentral: http://healthyliving.azcentral.com/importance-exercise-military-17575.html#
The Engineer. (17 November, 2015). Your questions answered: robots for extreme
environments.
Retrieved
31
August,
2016,
from
The
Engineer:
https://www.theengineer.co.uk/your-questions-answered-robots-for-extremeenvironments/
Weinberger, S. (4 April, 2012). Inside the navy's new autonomous robotics lab. Retrieved from
Popular
Mechanics:
http://www.popularmechanics.com/military/g832/inside-the-
navys-new-autonomous-robotics-lab/?
Yamazaki, J. N. (10 October, 2007). Hiroshima and Nagasaki death toll. Retrieved 4
September,
2016,
from
Children
http://www.aasc.ucla.edu/cab/200708230009.html
of
the
atomic
bomb: