We are just a couple week away from the Defense Advanced Research Projects Agency ( DARPA)Cyber Grand Challenge . The contest will take place August 4 in Las Vegas and pits the seven remaining finalist teams against each in a game of hacker capture the iris . Only there wo n’t be any dissident actively participating . Each team has built A.I. that will vie autonomously . It ’s part of an opening move from DARPA to develop new strategy for forestall cyber warfare .
But Tesla CEO Elon Musk is n’t into it . Late Thursday dark he tweeted that " It ’s all merriment & games until … " and then he linked to theSkynet Wikipedia page . That ’s the vicious , time - traveling A.I. net in theTerminatorseries that do out to kill all humanity . At a lower limit , there ’s a glossa - in - impudence aspect to Musk ’s tweet , but this is n’t the first time he ’s raised concerns about the destructive possibility of A.I. that turns against humans .
It sounds like something out of a scientific discipline - fiction film — and it ’s been the cannon fodder for countless — but these are real veneration there . The Cyber Grand Challenge started under the premise that humans simply are n’t able-bodied to ascertain , adapt , and respond tight enough to the develop cybersecurity headache as we find ourselves deeply and deeply into the " internet of things . " This challenge ask teams to create A.I. that learns and hunt out bugs itself , hoping that perchance someday the A.I. would be able to keep up with hack around the world .
Paramount Pictures
From one angle , that ’s a scary aspect and probably something real and worth considering . On the other hand , it ’s amazing that the human beings at the steering wheel of cybersecurity are able to keep up at all . DARPA write that set up a glitch once witness , " can take over a yr from first detection to the deployment of a solvent , by which time critical system may have already been violate . "
DARPA is looking for a cybersecurity system that would be " the first genesis of machines that can discover , prove , and fix package fault in actual - fourth dimension , without any assistance . " And while it ’s unlikely that Musk literally believe a Skynet - like techno - supervillain will go up from the ashes of this competition ( though , who knows ) , he does seem to find a duty to push the conversation on A.I. regularly , which serves an authoritative function itself as far as the public discourse on A.I. He ’s antecedently signedan open letterwarning about the potential threat of A.I. and , asInversenotes , has concern to it as our"greatest experiential threat . "
So , no , you do n’t need to start the search for Sarah Connor tonight , but Musk does n’t think the uninhibited forward motion of A.I. is something to take thinly . Maybe just practice groveling to our digital overlords a turn , that way whenithappens you ’ll be ready to do whatever it takes to survive .
