Pages

24 August 2017

Chinese Investment in AI Shows Scale of Cyber Ambition

by Steve King

While U.S. wrestles with civil discord, Beijing plows funds into dominating technologies of the future

If anyone doubts the intentions of the Chinese to achieve world dominance in cyberspace, look no further than their investment in artificial intelligence (AI). If their clear leadership in quantum computing isn’t enough evidence of their looming authority in what may be the final frontier, they have doubled down on developing the most advanced applications of AI on the planet.

Huge amounts of money are pouring into these technologies from China’s “private sector” investors, big internet companies, and its government, driven by the conviction that AI and quantum can remake their national security programs along with vast sectors of their economy.

Has it occurred to anyone yet that yelling at each other over the direction of our moral compass and playing chicken with a lunatic dictator-child characterizes a nation that probably doesn’t have its priorities in order?

Elon Musk Warns Of The Future Dangers Of Artificial Intelligence

China now has three significant advantages in this global arms race: an enormous pool of highly skilled and well-trained engineers and data scientists to work on the architecture and technology of cyberwarfare, a willing and legally unrestricted test data lake from 750 million internet users, and government leadership with political and financial support.

This huge data lake enables the country's engineers to train and test the algorithms necessary to manage events in cyberspace with increasingly less human intervention — and it's possible because China's citizens are not protected by privacy laws. This is a decided advantage over our own engineering efforts, which are forced to use much smaller data sets generated by simulations. Big data processing requires big data lakes. It should surprise no one that all of China's political leaders are engineers, and all of our own leaders are lawyers. Engineers like to create new and better ways of doing things. Lawyers do not.

China recently institutionalized the pursuit of AI and quantum into its national constitution. Published in July, the plan calls for the nation to be the leader in both of these technologies by 2030 — essentially a moon-shot declaration reminiscent of Kennedy's call on Congress to put a man in space within the decade. The Chinese government believes the AI industry alone will create $100 billion in economic activity and, combined with quantum, will result in a trillion-dollar economic boost and the dominance of cyberspace. China already has developed and implemented a closed internet based on quantum computing technologies. Soon, we will be paying for access.

Setting aside our leadership failures, our ambitions in this space are also burdened with data privacy laws that prevent us from aggregating the vast amounts of data required to adequately test and train these new systems. While DeepMind, Google's AI lab, has labored in courts for two years trying to get access to required medical records, X-rays and images, China's command-and-control economy, and its less restrictive privacy concerns, mean that the country can dispense video footage, medical records, banking information, and other vast springs of data whenever and to whomever it chooses. This political barrier makes it hard to compete.

The Chinese AI market is moving fast because people are willing to take risks and adopt new technology more quickly than their counterparts in America, underscoring the differences in mentality between engineering and law. I am not advocating a revision of our privacy laws. I am only pointing out that our choices have consequences.

On a pure numbers basis alone, China is producing more top engineers who are crafting AI algorithms for Chinese, not U.S. companies. Chinese universities and privately funded, government-supported firms are actively wooing AI researchers from across the globe, competing with U.S. companies, who are routinely now offering top researchers $500,000 annually, by doubling that compensation to a million dollars a year.

What's my point? We're losing the cyberwar.

We're losing on the educational front, as even after years of proof that there is a real and present and global threat in cybersecurity, of the 7,236 universities and colleges only 43 offer a BS degree in cybersecurity, and most of these are online. That's less than 1 percent.

We're losing on the economic front as we now spend at the rate of $90 billion a year, an 8 percent increase over last year, while the number of attacks increased 18 percent year over year to 400 raids every minute. While we spend hundreds of millions defending, cybercriminals are spending only $80 a day for a fully functioning attack kit.

We're losing on the informational front as recent brute-force attacks such as Wannacry and NotPetya clearly demonstrated. The attackers learned more about our defenses in a few minutes than we know even weeks later about the vectors the bad guys used in the attacks, or who they were or where they originated.

And we are losing on the technology front, as we have failed to marshal the resources necessary to apply AI, predictive analytics, machine learning and quantum techniques to our cybersecurity software or hardware in any useful way.

Most importantly, however, we are losing on the leadership front. We spend countless calories arguing over whether the far left or the far right is more responsible for violent protests, while China and North Korea and Iran and Israel invest heavily in the development of what they correctly understand will be the defining national defense and economic influences of the 21st century.

We can now watch in wonder as the Chinese launch program after program supported by billions of dollars in state-sponsored capital and underpinned by structured training and educational superiority, all aimed at advancing their preeminent position in cyberspace.

Steve King is the COO of Netswitch Technology Management.

No comments:

Post a Comment