Proliferation of Machine Intelligence capabilities in the commercial sector will provide low-cost access to intelligent systems, may spur unexpected threats.
Machine Intelligence (MI) is creating a new era of strategic technological competition that, like the space race for the previous generation, will define the balance of global power for decades. Policymakers around the world are taking notice. In September 2017, Russian president Vladimir Putin said, “Whoever becomes the leader in [MI] will become the ruler of the world.”
China’s New Generation of Artificial Intelligence Development Plan, released in 2017, echoes this view, saying, “Machine intelligence [is] the strategic technology that will lead in the future.”
Nations’ ability to produce and leverage MI technologies will define their economic competitiveness. Going forward, the global distribution of wealth, income, and soft power will be increasingly determined by nations’ ability to develop innovative MI applications and leverage those applications across their economies. In the United States, the tech sector directly accounts for 7.5 percent of GDP and provides over 7 million jobs, and is a key driver of growth and competitiveness across virtually every other sector. In sectors as diverse as manufacturing, health care, energy, retail, and advertising and media, deployment of MI technologies holds enormous promise for driving new growth.
Eight of the top 10 tech companies in the world are US-based, which allows the United States to exercise disproportionate power in cyberspace. US companies reach virtually every consumer in the world, helping us to spread our values around the world. Global supply chains enabled by technology and MI have driven global economic integration. Information and communication technology (ICT) infrastructure and digital inclusion are also key components of our efforts to build influence and support development and human rights in the developing world.
MI also has significant implications for hard power. Across the world, from the Chinese People’s Liberation Army (PLA) to the US Department of Defense (DoD), there is growing recognition that the next generation of military technologies will be driven by MI. The nature of conflict is shifting from today’s “informatised warfare” to “intelligentised warfare” or “algorithmic warfare”. Drone swarms, autonomous supply convoys, and remote medical care technologies have all seen significant investment from the world’s major militaries.
These changes will have profound impacts on the nature of conflict. Not only will military platforms and weapons change, but MI systems will fuel the development of entirely novel strategies, tactics, and concepts of operations. Perhaps the most transformative applications of military MI are in command and control (C2). MI-enabled C2 could develop entirely novel strategies, anticipate enemy tactics, accelerate intelligence, surveillance, and reconnaissance (ISR), and help coordinate activities of large numbers of dispersed units acting in tandem. As a greater share of decision-making on the battlefield happens at machine speed, human thinkers may be unable to keep up.
The focus of the policy debate around military applications of MI has centered around lethal autonomous weapons systems (LAWS). On one hand, many worry about the long-term implications of designing machines to kill human beings, citing the risks of losing control of the technology and of malicious actors being able to wreak havoc and create carnage at scale. On the other hand, some see an opportunity to free human soldiers from the most dangerous combat situations, help to alleviate the risks of human bias and misperception in combat, and for nations like Russia and China, an opportunity to upend the traditional military dominance of the United States by creating weapons that can fight and adapt at machine speed.
There is also growing concern MI will also empower non-state actors to develop new asymmetric warfare tactics. Proliferation of MI capabilities in the commercial sector will provide broad, low-cost access to intelligent systems. Just as cyber capabilities have enabled small groups to damage national security on a disproportionate scale, MI may spur new and unexpected threats. MI’s capacity to create indiscernible counterfeits of audio and video provide one example. In 2017, Adobe demonstrated a new product that, with 10 minutes of audio, can replicate a person’s voice exactly in limitless artificial audio.
The above is an extract from the report titled “A national machine intelligence strategy for the United States”, published this month by the Center for Strategic & International Studies, Washington D.C. Read the full report here.