Page 54 - The Economist Asia January 2018

SPECIAL REPORT
THE FUTURE OF WAR

2 new technologies have the potential not just to change the char-
acter of war but even possibly its supposedly immutable nature
as a contest of wills. For the first time, the human factors that
have defined success in war, “will, fear, decision-making and
even the human sparkofgenius, may be less evident,” he says.
Weapons with a limited degree of autonomy are not new.
In 1943 Germany produced a torpedo with an acoustic homing
device that helped it find its way to its target. Tomahawk cruise
missiles, once fired, can adjust theircourse usinga digital map of
Earth’s contours. Anti-missile systems are pre-programmed to
decide when to fire and engage an incoming target because the
human brain cannot react fast enough.
But the kinds of autonomy on the horizon are different. A
report by the Pentagon’s Defence Science Board in 2016 said that
“to be autonomous, a system must have the capability to inde-
pendentlycompose and selectamongdifferentcoursesof action
to accomplish goals based on its knowledge and understanding
ofthe world, itself, and the situation.” Whatdistinguishesauton-
omous systems from what may more accurately be described as
computerised automatic systems is that they work things out as
they go, making guesses about the best way to meet their targets
based on data input from sensors. In a paper for the Royal Insti-
tute ofInternational Affairsin London, MaryCummingsof Duke Effective—and expendable
University says that an autonomous system perceives the world
through its sensors and reconstructs it to give its computer
“brain” a model ofthe world which it can use to make decisions. alreadyclose to bringingaboutanotherrevolution. Robert Work,
The key to effective autonomous systems is “the fidelity of the the architect ofAmerica’s third offset strategy, stresses that this is
world model and the timeliness ofits updates”. not all about autonomous drones, important though they will
A distinction needs to be made between “narrow” AI, increasingly become. His main focus has been on human-mach-
which allows a machine to carry out a specific task much better ine collaboration to help humans make better decisions much
than a human could, and “general” AI, which has farbroader ap- faster, and “combatteaming”, usingunmanned and manned sys-
plications. Narrow AI is already in wide use for civilian tasks tems together.
such as search and translation, spam filters, autonomous vehi- Autonomous systems will draw on machine deep learning
cles, high-frequency stocktradingand chess-playingcomputers. to operate “at the speed oflight” where humans cannot respond
fast enough to events like cyber attacks, missiles flying at hyper-
Waiting for the singularity sonic speed orelectronic warfare. AI will also become evermore
General AI may still be at least 20 years off. A general AI important in big-data analytics. Military analysts are currently
machine should be able to carry out almost any intellectual task overwhelmed by the amount of data, especially video, being
thata human iscapable of. Itwill have the abilityto reason, plan, generated by surveillance drones and the monitoring of social-
solve problems, think abstractly and learn quickly from experi- media posts by terrorist groups. Before leaving the Pentagon, Mr
ence. The AlphaGo Zero machine which lastyearlearned to play Worksetup an algorithmic-warfare team to considerhow AI can
Go, the ancient strategy board game, was hailed as a major step help hunt Islamic State fighters in Syria and mobile missile
towards creating the kind of general-purpose algorithms that launchersin North Korea. Cyberwarfare, in particular, islikely to
will power truly intelligent machines. By playing millions of become a contest between algorithms as AI systems lookfornet-
games against itselfover 40 days it discovered strategies that hu- work vulnerabilities to attack, and counter-autonomy software
mans had developed over learns from attacks to design the best response.
thousands of years, and added In advanced human-machine combat teaming, UAVs will
some of its own that showed fly ahead of and alongside piloted aircraft such as the F-35. The
creativity and intuition. All over the place human pilot will give the UAV its general mission instructions
Mankind is still a long Worldwide spending on and define the goal, such as striking a particular target, but the
way from the “singularity”, the robotics, by sector, $bn UAV will be able to determine howitmeetsthatgoal by selecting
term coined by Vernor Vinge, a 70 from a predefined set of actions, and will respond to any unex-
science-fiction writer, for the Military pected challenges or opportunities. Or unmanned ground vehi-
moment when machines be- Industrial 60 cles might work alongside special forces equipped with wear-
come more intelligent than Commercial 50 able electronics and exoskeletons to provide machine strength
their creators. But the pos- Personal and protection. As Mr Work puts it: “Ten years from now, if the
sibility of killer robots can no 40 first through a breach isn’t a frickingrobot, shame on us.”
longer be dismissed. Stephen 30 Autonomous “uninhabited” vehicles, whether in the air,
Hawking, Elon Musk, Bill Gates on the ground orunderthe sea, offermanyadvantagesover their
and manyotherexpertsbelieve 20 manned equivalents. Apartfrom savingmoneyon staff, they can
that, handled badly, general AI often be bolder and more persistent than humans because they
could be an existential threat to 10 do notgettired, frightened, bored orangry. Theyare also likelyto
the human race. 0 be cheaper and smaller than manned versions because they do
In the meantime, military 2000 05 10 15 20* 25* not have to protect people from enemy attack, so they can be de-
applications of narrow AI are Source: Siemens *Forecast ployed in greaternumbers and in more dangerous situations. 1

14 The Economist January 27th 2018