The fact that Google is working directly for US Department of Defense has only been known for a few weeks. After thousands of employees of company protested against it, it now announced that it would not extend contract with Pentagon to 2019. Since 2017, Google had delivered artificial intelligence (KI) for Dassogenannte "project Maven" of Defense Ministry. The technology is designed to efficiently scan video footage of unarmed US surveillance drones for military-critical objects. This, according to FürchtenKritiker, could be beginning of a new kind of warfare in which machines and no more people decide who and what is a military goal – and could fight it independently.
The US security expert Paul WashingtonerThinktank, senior fellow at Center for a New American Security (CNAS), warns Indesdavor that AI researchers alone are giving generals question of how to militärtechnologischeingesetzt software. In his new book "Army of None: Autonomous Weaponsand The Future of War", Future of War was published in United States.
Online Time: HerrScharre, what is "project Maven" actually – and how important is it for US Department of Defense?
Paul Coulter: Project MAVEN is in ministry first before allemfür an innovation in procurement. Bureaucratic hurdles wurdendafür eliminated to be able not only with usual Rüstungsunternehmenzusammenarbeiten, but also with tech companies like Google – and as soon as possible new technologies such as artificial intelligence and cloud computing for military To make it usable. In case of project Maven, it is especially important to use available AI tools for image analysis in order to denriesigen amount of video material that drones deliver to US military. This is now a derartgroße mass that analysts are no longer able to follow. Alsosetzt now uses AI, which, thanks to machine learning, is getting better at recognizing and classifying objects. Artificial intelligence has been darinschon for years better than humans.
Online Time: WelcheRolle is Google playing project?
Crowd: First of all Googlestellt own Ki-software TensorFlowbereit. Which is freely available and an open source application, Imprinciple can be used by anyone. The US military does not employ m in weapons systems and certainly not in supposedly autonomous. But fact that Google is cooperating with US military has led to protests of employees Undletztlich that company will not extend its existing contract with DemVerteidigungsministerium.
Online Time: DieTech page Gizmodo has reported that at least two Anderekommerzielle companies are collaborating with project MAVEN, but re is no employee protests Hatman re. For Dasalte, Google's motto doesn't seem to apply: "Don't be evil."
Paul Coulter © Center for a New American securityCrowd: Ähnlichesließe are saying about companies that are currently applying for Einengroßen cloud computing mandate from Pentagon, as are EinigeTech corporations in race. Google may be a special case, due to its own history. The parent company Alphabet has, for example, purchased Ki company DeepMindversichert, whose technologies do not provide military. DeepMind is not involved in project MAVEN, but Dasalte pledge may have played a role now.
Online Time: is Esgrundsätzlich so that actually every form of Ki is a klassischeDual-use technology, so potentially militarily used equipment?
Crowd: The Ironiebesteht is that AI does not have to be malgroß adapted for military use, it can be easily built in. This must also be clear to engineers at Google. In case maven, free availability of Google's products means even that, in ory, terrorists could also download m. At same time, company opened its "Google AIChina Center" in Beijing a few months ago, although even Dienichtmilitärischen Ki plans of China would have to be problematic from point of view of Google Docheigentlich, for example with regard to denoset of biometric detection of People. Mirscheint that company situational its ethical standards in a variety of respects.
Date Of Update: 06 June 2018, 12:02
