人工智慧協助中國國防部開啟智慧防禦新領域

現代英語：

As one of the important representatives of the new round of scientific and technological revolution, artificial intelligence is the most cutting-edge topic in today’s scientific and technological field. AlphaGo Zero crushed its “AI predecessor” AlphaGo through self-learning, Baidu’s driverless car hit the road, and Apple’s mobile phone launched a new face recognition method… In recent years, the practical application of artificial intelligence has shown its huge driving force.

With the continuous advancement of artificial intelligence technology, how is artificial intelligence currently developing in the field of national defense? What role can artificial intelligence play in the field of national defense? How should artificial intelligence be developed in the future to better serve the field of national defense? Around these questions, the reporter interviewed Zhu Qichao, a researcher at the National University of Defense Technology.

Artificial intelligence has become a new focus of international competition——

Military powers are rushing to deploy

“From the perspective of the world situation, countries around the world, especially military powers, are rushing to deploy artificial intelligence. Government departments of the United States, Russia and other countries have all issued artificial intelligence-related strategies or plans, demonstrating that the country attaches great importance to artificial intelligence,” said Zhu Qichao.

Data shows that Russia’s “New Look Reform” that began in 2008 has made artificial intelligence a key investment area. In addition, Russia has also issued the “Concept of Developing a Military Science Complex by 2025”, emphasizing that artificial intelligence systems will become a key factor in determining the success or failure of future wars. In 2013, the European Union proposed a 10-year “Human Brain Project” to invest 1.2 billion euros in human brain research. In October 2016, the White House of the United States issued the “National Artificial Intelligence Research and Development Strategic Plan” to build an implementation framework for the development of artificial intelligence in the United States.

In Zhu Qichao’s view, many countries are promoting the development and application of artificial intelligence in the field of national defense. From the initial drones to intelligent information processing systems, bionic robots, etc., artificial intelligence has gradually penetrated into various fields of national defense and the military.

In recent years, the United States has used a large number of drones and logistics robots in the wars in Afghanistan and Iraq. Since 2014, the U.S. military has focused on investing in intelligent unmanned systems as a disruptive technology field of the “Third Offset Strategy”. In April last year, the U.S. Department of Defense announced the establishment of an algorithmic warfare cross-functional team to apply artificial intelligence to defense intelligence collection and analysis. According to reports, the U.S. Department of Defense recently officially ordered the establishment of a new artificial intelligence research center to integrate all artificial intelligence-related work of the Department of Defense.

Other countries are also accelerating their pace in this field and promoting the intelligentization of their armies. The Russian Military Industrial Committee plans to achieve 30% robotization of Russian military equipment by 2025, and its army’s wheeled and tracked ground combat robots have been deployed in the Syrian battlefield. South Korea and Israel have developed and used border patrol machines with automatic surveillance and autonomous firing capabilities. Israel has deployed highly autonomous “Harpy” drones in its territory. The South Korean Ministry of Defense also recently stated that it will invest 7.5 billion won by 2020 to promote the use of artificial intelligence in intelligence reconnaissance, command and control and other fields.

“It can be foreseen that various types of intelligent unmanned systems and combat platforms will be increasingly used on the ground, in the air, on the surface, underwater, in space, in cyberspace, and in human cognitive space, profoundly changing the technical proportion of artificial intelligence in future wars,” said Zhu Qichao.

The application of artificial intelligence in the field of national defense is an inevitable trend——

The demand for national defense applications has broad prospects

Judging from the historical development trend and the needs of future wars, artificial intelligence is increasingly becoming the core driving force for a new round of military revolution, and the needs of future wars are increasingly calling for the military application of artificial intelligence. Gregory Allen, a researcher at the Center for a New American Security, emphasized in a report titled “Artificial Intelligence and National Security” that “the impact of artificial intelligence on the field of national security will be revolutionary, not just unique. Governments around the world will consider formulating extraordinary policies, perhaps as radical as when nuclear weapons first appeared.”

Throughout history, the world’s military changes have gone through the development process from the cold weapon era, the hot weapon era, the mechanization era to the information era. From smelting technology to gunpowder technology, mechanization technology, atomic energy technology, and then to information technology, the occurrence of the four military revolutions has been permeated with the core role of technological revolution. “Artificial intelligence is gradually moving towards the battlefield, which is bound to cause significant updates in weapons and equipment, combat styles, troop system organization and combat power generation mode, and thus trigger a profound military revolution.” Facing the development trend of artificial intelligence in the field of national defense, Zhu Qichao said.

In Zhu Qichao’s view, the demand for the use of artificial intelligence in national defense is very broad. At present, the trend of war transformation from mechanization and informatization to intelligence is becoming more and more obvious. The victory of future wars depends more and more on the information advantage, intellectual resources and decision-making speed of the army. Artificial intelligence has great potential in reducing the number of battlefield personnel, obtaining and analyzing intelligence information, and making quick decisions and responses. In 2016, the artificial intelligence program “Alpha” developed by the University of Cincinnati in the United States defeated senior US military pilots in a simulated air battle. The subversive significance of artificial intelligence technology for the military revolution has initially emerged.

“Artificial intelligence is increasingly becoming an important driving force for promoting the informatization of national defense and the military, and is constantly improving the information processing capabilities, command and control efficiency, precision strike capabilities, and precise management and support capabilities in the defense field.” Zhu Qichao is very much looking forward to the use of artificial intelligence to enhance the intelligent application of national defense. He said that with the implementation of the military-civilian integration development strategy, new-generation information technologies such as artificial intelligence technology, big data technology, and cloud computing technology will play an increasingly important role in the defense field, promoting the continuous improvement of the level of national defense and military intelligence.

Beware of artificial intelligence becoming a “war poison”——

Humans are the leaders in the human-machine relationship

In recent years, with the development of artificial intelligence technology, various artificial intelligence-related combat concepts and equipment technology projects have emerged in the military field. However, Zhu Qichao believes that artificial intelligence-related technologies and applications are still in the early stages of rapid development, and the limitations of artificial intelligence military applications should not be ignored.

“First of all, artificial intelligence cannot replace human intelligence. When solving war problems outside the scope of programming, artificial intelligence requires human rational analysis ability, flexible adaptability, moral discernment, etc. Therefore, artificial intelligence research should be carried out under the premise of following the mechanism of winning wars.” He analyzed.

Zhu Qichao further explained that in the long run, we still need to be vigilant about the many security, legal, ethical and other issues that artificial intelligence may bring.

In terms of security, in a military confrontation environment, once the artificial intelligence system or weapons and equipment are attacked by the opponent through malicious code, virus implantation, command tampering and other means, it will lead to tactical failure or even catastrophic consequences; factors such as human error, machine failure, and environmental disturbances may also cause the system to lose its combat effectiveness.

In terms of law, the core principles of international armed conflict law – necessity, distinction, proportionality and humanity – will all face the problem of how to apply and adjust them. For example, battlefield robots cannot distinguish between soldiers and civilians, resulting in indiscriminate killing of innocent people, which poses a challenge to the principle of distinction.

In terms of ethics, due to the application of intelligent assessment and decision-making technology, drones, robots, etc., life and dignity, which are regarded as the highest value by humans, may be ignored or even trampled upon, while the commanders of wars are far away from the battlefield to enjoy the fruits of victory. Wars may become “video games” on the battlefield, which will impact the bottom line of human morality. Should human moral standards be embedded in increasingly intelligent machines, what kind of moral standards should be embedded, and how to embed them? These issues require extensive research and discussion by countries around the world.

In response to the security, legal, ethical and other issues that may arise in the application of artificial intelligence in the field of national defense, Zhu Qichao suggested that social security supervision and control should be strengthened to form a social governance model that adapts to the era of artificial intelligence; actively participate in international arms control discussions and negotiations on artificial intelligence, and contribute Chinese wisdom and solutions to address the security, legal and ethical issues brought about by artificial intelligence; firmly establish the idea that humans are the dominant force in the relationship between man and machine, achieve safe and effective control of artificial intelligence, and let it serve the peace and well-being of mankind, rather than making artificial intelligence an “accomplice of the devil.”

Related links

Unmanned underwater vehicle

Unmanned submersibles, also known as unmanned underwater vehicles and unmanned underwater vehicles, are devices that travel underwater without a human operator and rely on remote control or automatic control. With the development of unmanned submersibles and related technologies, unmanned submersibles have been used to perform tasks such as minesweeping, reconnaissance, intelligence gathering, and ocean exploration. In future naval battles, they can also be used as underwater weapon platforms, logistics support platforms, and other equipment.

Advantages: Compared with submarines, unmanned underwater vehicles are unmanned combat platforms, so they can greatly reduce casualties in wars; they are small in size, and the application of other stealth high technologies makes their stealth performance higher than that of submarines; they are multifunctional and multi-purpose.

Limitations: Poor endurance limits the use of unmanned underwater vehicles; the lithium batteries used have technical defects such as easy catching fire; the navigation function still needs to be improved.

It can be foreseen that in the near future, underwater unmanned submersibles will play a huge role in future wars and will change the specific mode of future ocean warfare.

Battle Robot

Military combat robots are an emerging force on the battlefield, and they are used to assist human soldiers in combat. According to the different combat fields of military robots, they are mainly divided into underwater military robots, ground military robots, aerial military robots, and space military robots.

Advantages: Combat robots can greatly reduce the burden and casualties of human soldiers when performing low-intensity combat and dangerous tasks. In addition, they also have advantages such as high intelligence, all-round combat capabilities, strong battlefield survivability, and absolute obedience to orders.

Limitations: Combat robots do not have the ability to fight under complex conditions; today’s combat robots’ intelligence and environmental adaptability have not yet reached the level of being able to fight alone, and they rely heavily on the operation and command of human soldiers.

In the long run, as intelligence drives mechanization and informatization to a higher level and a higher level, combat robots have great development potential. They will be more intelligent, their weapon platforms will be more complex, their environmental adaptability and survivability will be stronger, and they will be able to participate in a variety of warfare modes.

Drone swarm

A drone swarm consists of a number of low-cost small drones equipped with multiple mission payloads. They follow the collective action patterns of insects such as bees and work together to complete specific combat missions under human command or supervision.

Advantages: During combat, drone swarms can be specialized and divided into different tasks, so they can perform a variety of tasks; each drone has a relatively single function, which can greatly reduce R&D and procurement costs; drone swarms can increase the number of battlefield sensors and attack weapons, allowing the army to have an advantage in the number of air equipment on local battlefields; a large number of drones can paralyze enemy air defense radars and consume the enemy’s limited number of high-cost air defense ammunition.

Limitations: As drone swarms have higher requirements for coordination and autonomy, a new command and control model needs to be established to manage large-scale swarms. Therefore, it faces the challenges of mastering key technologies such as collaborative combat algorithms, communication between swarm individuals, and remote command and control.

In the future, drone swarms will drive future air combat equipment to present characteristics such as cheaper airframes, autonomous platforms, and smaller payloads, which may have a revolutionary impact on the development ideas of future aviation equipment systems.

現代國語：

作為新一輪科技革命的重要代表之一，人工智慧是當今科技領域最前線的課題。 AlphaGo Zero透過自我學習碾壓「AI前輩」AlphaGo、百度無人汽車上路、蘋果手機開啟新的刷臉認證方式…近年來，人工智慧的實際應用顯示其技術巨大的驅動力。

在人工智慧技術不斷進步的背景下，人工智慧在國防領域目前發展如何？人工智慧在國防領域能發揮什麼作用？未來應如何發展人工智慧使其更好地服務國防領域？圍繞著這些問題，記者採訪了國防科技大學研究員朱啟超。

人工智慧成為國際競爭新焦點——

軍事強國紛紛搶灘部署

「從世界局勢來看，世界各國尤其是軍事強國都在搶先佈局人工智慧，美、俄等國家政府部門均發布了人工智慧相關戰略或規劃，彰顯國家層面對人工智慧的高度重視。」朱啟超表示。

資料顯示，俄羅斯始於2008年的「新面貌改革」將人工智慧作為重點投資領域。此外，俄羅斯也發布《2025年前發展軍事科學綜合體構想》，強調人工智慧系統將成為決定未來戰爭成敗的關鍵要素。歐盟在2013年提出為​​期10年的“人腦計畫”，擬斥資12億歐元進行人類大腦研究。 2016年10月，美國白宮發布《國家人工智慧研究與發展策略規劃》，建構美國人工智慧發展的實施架構。

在朱啟超看來，不少國家都在推動人工智慧在國防領域的發展運用，從最初的無人機到智慧化資訊處理系統、仿生機器人等，人工智慧逐步滲透到國防和軍隊各個領域。

近年來，美國曾在阿富汗戰爭、伊拉克戰爭中大量運用無人機和後勤作業機器人。自2014年以來，美軍已將智慧化無人系統作為「第三次抵銷戰略」的顛覆性技術領域給予重點投資。去年4月，美國國防部宣布成立演算法戰跨職能小組，旨在將人工智慧用於國防情報蒐集和分析領域。據報道，日前美國國防部正式下令成立一個新的人工智慧研究中心，整合國防部所有的人工智慧相關工作。

其他國家也在這個領域加快步伐，推動軍隊智慧化建設。俄羅斯軍事工業委員會計畫在2025年之前實現俄軍裝備30%的機器人化，其軍隊輪式和履帶式地面作戰機器人已經投入敘利亞戰場。韓國和以色列開發和使用具有自動監視和自主決定開火能力的邊境巡邏機器，以色列已在其境內部署自主性很高的「哈比」無人機，韓國國防部也在近期表示將在2020年之前投入75億韓元用於推動人工智慧在情報偵察、指揮控制等領域的運用。

「可以預見，各類智慧化無人系統與作戰平台將在地面、空中、水面、水下、太空、網路空間以及人的認知空間獲得越來越多的應用，深刻改變著未來戰爭人工智慧的技術比重。」朱啟超說。

人工智慧運用於國防領域是大勢所趨——

國防運用需求前景廣闊

從歷史發展趨勢和未來戰爭需求來看，人工智慧越來越成為推動新一輪軍事革命的核心驅動力，未來戰爭需求也越來越呼喚人工智慧的軍事應用。新美國安全中心研究員格雷戈里·艾倫在其主筆的一份題為《人工智能與國家安全》的報告中強調：“人工智能對國家安全領域帶來的影響將是革命性的，而不僅僅是與眾不同的。世界各國政府將會考慮制定非凡的政策，可能會像核武器剛出現時一樣徹底。”

縱觀歷史，世界歷次軍事變革經歷了從冷兵器時代、熱兵器時代、機械化時代到資訊化時代的發展歷程，從冶煉技術到火藥技術、機械化技術、原子能技術，再到資訊技術，四次軍事革命的發生都貫穿著技術革命的核心作用。 「人工智慧逐步走向戰場，勢必會引起武器裝備、作戰樣式、部隊體制編制和戰鬥力生成模式顯著更新，進而引發一場深刻的軍事革命。」面對人工智慧在國防領域的發展態勢，朱啟超表示。

在朱啟超看來，人工智慧的國防運用需求非常廣闊。當下，戰爭形態由機械化、資訊化轉型為智慧化的趨勢愈發明顯，奪取未來戰爭的勝利越來越取決於軍隊的資訊優勢、智力資源和決策速度。而人工智慧在減少戰場人員數量、獲取和分析情報資訊、快速決策和反應等方面具有巨大的潛力。 2016年，美國辛辛那提大學研發的人工智慧程式「阿爾法」在模擬空戰中擊敗了美軍資深飛行員，人工智慧技術對於軍事革命的顛覆性意義已初步顯現。

「人工智慧越來越成為推動國防和軍事資訊化建設的重要驅動力，不斷提升國防領域的資訊處理能力、指揮控制效率、精確打擊能力和精準管理保障能力。」朱啟超對人工智慧提升國防領域智慧化運用非常期待，他表示，隨著軍民融合發展戰略的實施推進，人工智慧技術、大企業數據將不斷提昇軍事化數據等新一代資訊技術將越來越重要在國防領域推動國防和電力提升。

警惕人工智慧成為「戰爭毒藥」——

人類是人機關係主導者

近年來，隨著人工智慧技術的發展，軍事領域湧現出各種人工智慧相關作戰概念和裝備技術項目，但朱啟超認為，目前人工智慧相關技術與應用還處於快速發展的初級階段，不應忽視人工智慧軍事應用的限制。

「首先，人工智慧並不能取代人類智慧。人工智慧在解決可程式範圍外的戰爭問題時，需要人類的理性分析能力、靈活應變能力、道德分辨能力等，因此，要在遵循戰爭制勝機理的前提下進行人工智慧研究。」他分析道。

朱啟超進一步說明，長期來看，還需要警惕人工智慧可能帶來的安全、法律、倫理等諸多問題。

安全方面，軍事對抗環境下，人工智慧系統或武器裝備一旦被對手透過惡意程式碼、病毒植入、指令篡改等手段攻擊，將帶來戰術失利甚至災難性後果；人為錯誤、機器故障、環境擾動等因素也可能使得系統失去戰鬥效力。

在法律方面，國際武裝衝突法中的核心原則——必要性、區別性、相稱性和人道性都將面臨如何適用和調整的問題。例如，戰場機器人無法區分軍人與平民而造成濫殺無辜給區別性原則構成挑戰。

倫理方面，由於智能化評估決策技術、無人機、機器人等的應用，人類奉為最高價值的生命和尊嚴可能受到漠視甚至踐踏，而戰爭的指揮者卻遠離戰場享受戰爭勝利的果實，戰爭或將成為搬上戰場的“電子遊戲”，這將衝擊人類的道德底線。是否應該將人類的道德標準嵌入日益智慧化的機器、嵌入什麼樣的道德標準、如何嵌入？這些問題需要世界各國的廣泛研究與探討。

針對人工智慧在國防領域運用過程中可能出現的安全、法律、倫理等問題，朱啟超建議，應加強社會安全監督管控，形成適應人工智能時代的社會治理模式；積極參與人工智能國際軍備控制討論與談判，為應對人工智能帶來的安全、法律與倫理問題貢獻中國智能和中國；牢固幫助

相關連結

無人潛航器

無人潛航器，也可稱為無人水下航行器和無人水下運載器等，是沒有人駕駛、靠遙控或自動控制在水下航行的器具。隨著無人潛航器及相關技術的發展，無人潛航器已被用於執行掃雷、偵察、情報蒐集及海洋探測等任務，在未來海戰中還可作為水下武器平台、後勤支援平台等裝備使用。

優點：與潛水艇相比，無人潛航器是無人作戰平台，因此可以大大降低戰爭的傷亡；體形小，加上其他隱身高科技的應用使其隱身性能高於潛艇；多功能，多用途。

限制：續航性差限制無人潛航器使用範圍；所用鋰電池有易著火等技術缺陷；目前導航功能尚需完善。

可以預見，在不久的將來，水下無人潛航器必將在未來戰爭中發揮巨大作用，並將改變未來海洋作戰的具體模式。

戰鬥機器人

軍用戰鬥機器人作為戰場上的新興力量，是配合人類士兵作戰的角色。依軍用機器人作戰領域不同主要分為水下軍用機器人、地面軍用機器人、空中軍用機器人和太空軍用機器人等。

優點：戰鬥機器人在執行低強度作戰和危險任務時可以大大減輕人類士兵的負擔和傷亡。此外，其還具有較高智能、全方位作戰能力、較強戰場生存能力、絕對服從命令等優勢。

限制：戰鬥機器人不具備複雜條件下的作戰能力；如今戰鬥機器人的智慧化和環境適應能力還未達到單獨作戰程度，很大程度依賴人類士兵的操作和指揮。

從長遠來看，隨著智慧化牽引機械化和資訊化向更高層次、更高層次發展，戰鬥機器人發展潛力巨大，其智慧化程度將更高、武器平台將更複雜、環境適應和生存能力也將更強，能夠參與的戰爭模式也將多種多樣。

無人機蜂群

無人機蜂群由若干配備多種任務負荷的低成本小型無人機組成，它們參考蜜蜂等昆蟲的集體行動模式，在人類指揮或監管下共同完成特定作戰任務。

優點：作戰時無人機蜂群可專業化分工，因此能執行多種任務；每架無人機功能相對單一，可大幅降低研發和採購成本；無人機蜂群可增加戰場感測器和攻擊武器數量，使軍隊在局部戰場擁有空中裝備數量優勢；大量無人機可癱瘓敵人防空雷達，消耗敵人有限數量的高成本防空彈藥。

限制：由於無人機蜂群對協同和自主的要求更高，需要建立管理大規模蜂群的全新指揮控制模式，因此面臨攻克協同作戰演算法、群集個體間通訊、遠端指揮控制等關鍵技術的挑戰。

未來，無人機蜂群將牽引未來空中作戰裝備呈現機體廉價化、平台自主化、載重小型化等特點，可能對未來航空裝備體系的發展思維產生變革性影響。

中國國防報記者 潘 娣 通訊員 孫 清 高旭堯

中國軍網 國防部網
2018年7月11日 星期三

中國原創軍事資源：http://www.81.cn/jfjbmap/content/2018-07/11/content_210708888.htm