The Defense Advanced Research Projects Agency recently announced the Cyber Grand Challenge, a seven-team, $3.75 million-prize-pool hacking tournament that concludes in Las Vegas this August. DARPA’s stated goal is to find new strategies for countering cyberwarfare, but Elon Musk isn’t buying it. In fact, he think the Cyber Grand Challenge might just leave us with something like Skynet, the hostile A.I. that routinely tries to destroy humanity in the Terminator franchise. Musk tweeted a not-so-veiled warning on Thursday morning — and it’s not the first time he’s gone on the record as being fearful of a malicious supercomputer.
On paper, DARPA wants to build an automated artificial intelligence that is capable of detecting and resolving bugs in a computer security system. Essentially, they want to create an unsupervised, autonomous A.I. hacker extraordinaire that will detect system vulnerabilities and patch them itself. As it stands today, of course, people perform the often thankless duties of cybersecurity. Expert hackers are adept at finding and fixing susceptibilities, but as cyberwarfare becomes more prevalent their demand for their skills could surpass the supply. The process of fixing a flaw, DARPA writes, “can take over a year from first detection to the deployment of a solution, by which time critical systems may have already been breached.”
And, the demand for quick fixes to ever-present security issues continues to rise as more and more everyday devices communicate information over the internet. DARPA is arguing that a cybersecurity A.I. system would be “the first generation of machines that can discover, prove and fix software flaws in real-time, without any assistance,” making the whole world more secure.
Or, y’know, it could lead to the systematic destruction of humanity at the cold metal hands of the machines.
Musk’s tweet may seem like it’s in jest, but he has a long record of being a watchdog toward the dangers of super-powerful artificial intelligence. He signed an open letter warning about A.I. and has called the technology our “greatest existential threat.”
Musk knows that this DARPA challenge probably won’t spawn Skynet, but it seems important to him that people consider the possibility of malevolent artificial intelligence. Once we as a society unleash such a beast, doomsayers say, there’ll be no holding it back.