White House no longer wants to rely on Google & IBM to weaponize AI

Plans to increase the AI and quantum computing budget shows that the Pentagon will no longer be a backseat driver in the development of cutting-edge military tech. Expect US capabilities to grow rapidly and in greater secrecy.

The announcement last week that more money of the Trump administration’s proposed $4.8 trillion budget would be earmarked for the Department of Defense and the National Science Foundation is a significant step in the ongoing Artificial Intelligence arms race. 

Fearful that the US is falling behind China in this race, many technologists have urged the Trump administration to back research on AI-based weapons and transportation and on quantum computing, which harnesses quantum mechanics to reach potentially vastly faster processing speeds and stronger encryption capabilities than existing supercomputers

The new budget proposals are significant mainly because they reveal that the US government now recognizes that the battle over the weaponization of AI centers around a battle for human talent. 

Until now, much of the US’s AI capability has been an offshoot of the private endeavours of tech giants like Google, Amazon, Microsoft and IBM. 

The problem is not simply that US universities and government labs have lost much of their talent to these businesses. They have also lost them to other countries. Much of the critical research being done is now taking place in tech hotspots like Toronto, London and Beijing. 

And there is an even potentially bigger ethical dilemma: many leading pioneers of AI, like Demis Hassabis at Google DeepMind and Elon Musk at the US rocket company SpaceX are among more than 2,400 signatories to a pledge which opposes military firms and nations from building lethal autonomous weapon systems. 

These developments mean the US government has a significant battle ahead to win the hearts and minds of future AI and quantum computing talent. 

Which is why the detail of the proposed increased budget is so revealing: the Defense Advanced Research Projects Agency, a research arm of the US Department of Defense, will increase funding of AI-related research to $249 million from $50 million, while the National Science Foundation’s present funding of about $500 million will increase to $850 million, with $50 million being earmarked specifically to help train AI experts.

The machines will not know what they are doing. But governments and the well-remunerated professors and Tesla engineers recruited by the Pentagon certainly will. And so too will Beijing, whose budgets and capabilities will ramp up accordingly.