NORBERT BIEDRZYCKI
4 min readPressRead this in Polish

Military robots may decide whether we live or die

Military robots may decide whether we live or die

My article published in Business Insider on 2 March 2019.

The military is constantly testing electronic command and training systems, object recognition techniques and battlefield management. What is more, in the near future decisions to deploy offensive weapons may be made directly by algorithms. In the case of defensive weapons, decisions to open fire or to use combat systems are very often already made by fully autonomous machines, without human involvement (for example, the defence systems of a tank or a combat helicopter).

Politicians, experts and representatives of the IT industry increasingly realise that the problem of autonomous weapons is not an abstraction. For several years, broader and stronger initiatives have been raising these issues. The United States should “commit to negotiating a binding treaty that sets limits on the autonomy of weapons systems. It should prevent the development, production and use of fully autonomous weapons”, says Mary Wareham of Human Rights Watch. The activist and film-maker regularly monitors the issue as part of an international coalition, the “Campaign to Stop Killer Robots”.

The British non-governmental organisation Article 36, which pays a great deal of attention to autonomous weapons, argues that political controls on armaments should be regular and based on publicly available, transparent analyses. Both organisations put a lot of energy into developing clear definitions of autonomous weapons. Signatories of international petitions try to reach politicians and present their point of view at international conferences.

One of the most recent international initiatives is this year’s pledge organised by the Boston-based Future of Life Institute. 160 AI-related companies in 36 countries and 2,400 individuals signed a declaration stating that “lethal autonomous weapons pose a clear threat to the citizens of every country in the world” and that they will not participate in their development. The signatories included Demis Hassabis, Stuart Russell, Yoshua Bengio, Anca Dragan, Toby Walsh and Elon Musk, the founder of Tesla and SpaceX.

Regulation is needed, but utopian

Initiatives of this kind face a difficult task. First, monitoring military projects is complicated by its very nature. Until an open international conflict reveals which technologies are actually in use, tracking which weapons are being developed, researched and deployed is almost impossible.

Another obstacle to clear, binding norms and agreements is the nature of algorithms. It turns out that we are very used to seeing weapons as material objects (whose use may or may not be banned). When we talk about software, algorithms, neural networks and, even more so, artificial intelligence, legal regulation struggles with the intangible character of these tools.

There is also the problem of blurred responsibility, which I have written about in my texts on autonomous cars. If a tragedy happens, who is responsible for it? The programmer who wrote the algorithms so that they could make their own choices? Or the person who trained the neural networks?

Military robots and death, Norbert Biedrzycki blog, Business Insider

The military says it is about reconnaissance

The military, however, responds that instead of banning, we should encourage innovations that reduce civilian casualties – or so say officers lobbying for the most advanced projects. The point, they argue, is not for algorithms to destroy the enemy and the civilian population. They are above all meant to help assess the situation on the battlefield and analyse it for opportunities to gain a tactical advantage. In this view, intelligent technologies mainly lead to more efficient data processing, which then makes it easier to identify targets and reduces accidental civilian casualties.

Let us not count on human rights

Algorithms unleashed on the battlefield could lead to tragedies of unprecedented scale. Toby Walsh, a professor of artificial intelligence at the University of New South Wales in Australia, said: “Autonomous weapons will not respect human rights. Machines may roam the battlefield, on the ground or in the air, choosing their own targets and then destroying them with fire or missiles, without human involvement.”

Artificial intelligence has enormous potential to help humanity. Unfortunately, it also has a dark side. We can only hope that politicians and military leaders have enough imagination to see in time all the consequences of the spread of autonomous weapons.

The full article (in Polish) can be read at this link.

Related articles:

– A machine will not hug you, but it may listen and offer advice

– Can machines tell good from evil?

– Only God can count as fast: the world of quantum computers

– Machine learning. Computers are no longer babies

– How machines reason

Read next

  1. BUSINESS INSIDER: How machines think

    My article in Business Insider published 22 October 2016. Technological advances make you wonder: is this the start of a new world ruled by intelligent machines? Is there such thing as artificial intelligence? The development of artificial intelligence in robotics is not as unlikely as you might think. A

  2. The soulless logic of algorithms seduces business and flirts with us, the customers

  3. Machine Ecology

    Machine Ecology. The earth is sick, and the causes of its disease lie in our choices, our myopia, and our lack of imagination.