Terminator's Skynet is Coming!

"Hello, Internet. Welcome to — woah, started a bit early there. Don't worry, though, I'll be back. Just call me the Matinator; the, the TerminatPat, the Terminat —"

- MatPat

"Hello, Internet. Welcome to FILM THEORY! Have you guys ever seen Terminator? Is this audience old enough to appreciate a good, burly Austrian accent? Who else out there would look back on Terminator 2's slowly-sinking-into-lava-thumbs-up and give it two slowly-sinking-into-lava-thumbs-ups."

- MatPat

Terminator's Skynet is Coming! (subtitled DEATH by A.I.!) is the 17th episode of Film Theory and its one-off episode on Terminator on The Film Theorists.

Description
*****Thanks to Audible for sponsoring this video and allowing us to deliver more videos FASTER to the Loyal Theorists!*****

Transcript
Regardless of some of the turns the franchise has had over the years, one thing you have to admit is how awesome a job the movies do at creating a dark post apocalypse where robots have taken over, and humans fight for their very survival. It's the kind of scenario that gets the old theorist cogs a turning. Is this kind of world just the stuff of nightmares, or is the world of the Terminator our great big beautiful, machine infested, post-apocalyptic, tomorrow? Could a world in which all of humanity is enslaved to an all-powerful supercomputer like Terminator’s Skynet ever possibly become a reality? Should we be worried about evil killer robots who look and act like people? Believe me, Watson already creeps me out a bit. Did you see him decimate the competition on Jeopardy? In the original Terminator, our protagonist Kyle Reese described a future where the infamous Skynet came online and destroyed the world in 1997. Terminator 3 portrayed a similar armageddon, taking place in 2004. Today, we're all still around, killer cyborg men with dare I say excellent dental work aren't patrolling the streets and all the color in the world hasn't become that weird grayish blue. However, the newest film Terminator Genisys predicts that the take over from our robotic overlords will occur in as little as two years: 2017. Now the other films got their dates wrong, so this one couldn't possibly be right either right? I mean, it couldn't be right, we're all not gonna die in two years are we? Let's just take a look and find out the truth behind the story. First, what exactly would need to happen for a rogue AI like Skynet to take over the world? Well, at its core, Terminator One’s apocalypse happens because Skynet- a program designed specifically for military use-revolts against its creators and launches all of the US's nukes, killing billions in the blink of an eye. It's then able to take control of all military machinery jacked into its systems, and use it against the surviving human resistance. Now, you might think we're quite a few years away from making that kind of blanket military strike, but it turns out that weaponized machines are reality, and this sort of robopocalypse is actually a lot closer than it was when the first few Terminator films were being made. If you don't believe me that this is a big deal, take a listen to what some of the world's smartest scientific minds are saying about advances in weaponized tech. “We could one day lose control of AI systems via the rise of super intelligences that do not act in accordance with human wishes, and that such powerful systems would threaten humanity”. This line doesn't come from a sci-fi movie, it's a real letter presented at the International Joint Conferences on artificial intelligence, in January of this year [2015]. Signed by Stephen Hawking, Elon Musk of SpaceX and Tesla Fame, Steve Wozniak, the Apple guy who isn't Steve Jobs, and many many many more. These science and tech geniuses are genuinely worried about killer machines taking over in the future, and you know what, maybe we should be too. Right now, our killer robots are just called drones. The kind that can parachute in your latest dating advice book from amazon.com, and they're quickly becoming just another one of those technology things we take for granted in our everyday lives. But drones have been around for a lot longer than they've been delivering bulk bags of cheez-its to your doorstep. In fact, they were originally developed in the 70s and the first weaponized drone was used in war by Iran in the late 1980s during the Iran-Iraq war, and the United States invested heavily in unmanned aerial vehicles or UAVs for all you COD heads, during the first Gulf War in the 90s. The program was incredibly successful, which led to the U.S. to up its production of drones for use in war, most recently armed drone strikes have been used heavily in the war on terrorism across the Middle East. As of June 2015, the total death toll from US drone strikes was estimated to exceed 6,000, with up to 890 civilians and about 175 children killed in Pakistan alone. That's a lot of innocents killed, and these sorts of drones aren't just being rolled out in the Middle East. Drones are becoming more and more common in North America as well. Several law enforcement agencies across the U.S. have been using drones for surveillance, and many bureaus have been requesting permission to equip their UAVs with tear gas and guns that fire rubber bullets. Non-lethal weaponry to be sure, but still, weaponry. In the hands of computer-controlled devices. And that starts to get scary. Now you might be thinking, ‘yeah it's a little creepy I guess, but they’re still our drones, we have control over them’. Well, you might want to think again, because things can get hacked. We previously talked about how easily hackers can get into everything we own- from our phones to our computers to our cars, in our Watch_Dogs episodes over on Game Theory, and drones are no exception. In 2011, a CIA stealth drone was hijacked by a group of Iranians, who not only brought it down, but did it in a way that they could reverse-engineer the technology for themselves. And the craziest part of this? The instructions on how they did this can be found with a quick google search. Now, we here at Theorists Inc. don't condone this sort of behavior, so we're gonna elect not to share how to find this information, but for those of you smart enough to use related keywords and know how to click a button with a giant magnifying glass on it, you'll find yourself a PDF explaining exactly how to generate fake signals that can fool the drones GPS receivers, as well as how to replace legitimate signals with fake ones. Can you imagine if a rogue AI program like Skynet were able to get into systems of drones like this? It's a wild hypothetical scenario, right? Hypothetical might not be the right word to describe it. In early July, a massive vulnerability was found in Adobe Flash. You know, the program that allows you to play slime volleyball on addictinggames.com, which allowed hackers to install malware on thousands of systems without users even realizing anything was wrong. Later that same month, Microsoft rushed out an emergency patch to fix a ginormous security bug that allowed hackers to remotely execute malicious code on computers running any supported version of Windows. Two gaping security holes that affect pretty much anyone who has ever touched a computer ever, but you know what the scariest part about this is? Several parts of the US government, including branches as large and powerful as the Navy, currently use Windows as their operating system. If a rogue AI found a similar hole and simultaneously exploited it on every system it could, that's starting to sound a lot like Skynet, isn’t it? So, we've managed to lay the foundation, or at the very least the major building blocks for a Skynet-esque Doom program to take over. Our computer systems are imperfect and vulnerable, and our military and police forces are operated by machines more and more every day. So what's missing? Why aren't we the loyal subjects of an all-powerful Robo overlord already? Well, the answer to that question is true artificial intelligence. In all of the films, Skynet is a self-aware system. That's what made it so dangerous. It was able to comprehend things like life and death, the fact that it existed, and even had a self-preservation instinct. That's why it launched all of those nukes and killed humans in the first place. It saw us humans as a threat to its existence, and it got scared. And if all of that sounds to you like the plot of Avengers Age of Ultron, you're dead on. Now obviously no technology like this currently exists in our world, but believe it or not, it's not entirely out of the realm of possibility. In fact it's a subject that's been seriously speculated on by scientists and science fiction writers alike. Really there's only one thing standing between us and true sentient AI like Skynet, and that thing is time. Eventually, it will happen, in an event known as the singularity. Now I'm sure a lot of you have heard that word tossed around in the past, but let's define it here. The technological singularity will be a point in time where technology becomes so powerful, and the exponential advancement of science will be so quick, that a super intelligence is created. Singularity itself is a term borrowed from astrophysics. It refers to a point in space-time where normal physics no longer apply. Think the center of a black hole, where there's so much mass that time ceases to be a factor. Now the singularity could be achieved in one of four ways, but to save us all time, let's just talk about the one that Skynet fits into: Where computer network becomes so advanced that it becomes self-aware. If that were to ever happen, it would basically be impossible for us humans to keep up. A computer that was aware of itself could use all of its intelligence to design an even better computer, which would then design a better computer which would then design a better computer and so on and so forth. It would get to the point where a super computer could design an even better supercomputer in milliseconds, and if it decided to turn against us, like Skynet did, well, we probably wouldn't stand a chance. So, why are we talking about this now? Again this sounds like something out of science fiction. There is no way a computer could get so advanced so quickly, and it especially isn't feasible in the next few years, right? Right? Technology is fickle, due to the nature of electronics, Computer Sciences advanced at a much faster rate than other fields. This is so well documented that there's even a neat little name for the concept. It's called Moore's Law, and it states that the number of transistors- basically computer thinky bits- on a circuit board, doubles every two years. Now, this seems like a very simple and very rough estimation, but let me tell you this is held true for decades, and has only slowed down recently. In 2015, representatives at Intel stated that “our cadence today is slower to two and a half years than two”. Still, that is incredibly quick for the type of advancements we're making. In layman's terms, this means that computers are getting really powerful, really quickly, and if you want proof of those advancements just look at the landscape of computers in the past decades. Forty years ago, in the 70s, a computer could take up an entire room in a laboratory. 30 years ago computers were almost non-existent in the general market, and their screens could only display text or rudimentary drawings pixel art. Twenty years ago, in the 90s, while a bit more common, computer systems weren't very powerful. They were mostly used for office work or to play really unadvanced video games. Ten years ago, in the 00s, computers were now common, but garbage compared to today's standards. I remember once having to buy a thumb drive for school, and 156 megabit drive cost right around $50. $50! Today, a drive with that little memory space literally wouldn't be worth the plastic and metal used to create it. The phones that many of us have in our pockets today have more power than what was used to go to the moon in the 60s, and our very best computers can render three-dimensional imagery so realistic, that you might be unable to distinguish it from the real thing. And for as simple as it may seem, models created using Moore's Law outlined all of this, and some scientists have been using it to predict future trends in technology. One such man is named Ray Kurzweil, computer scientist, inventor, futurist, and director of engineering at Google. Kurzweil is the definition of a genius and a pioneer of computing technology, having been heavily involved with optical character recognition, text to speech, and speech recognition technology in the past. He was also named one of PBS's 16 revolutionaries who made America. Read his bio, it's very clear he knows what he's talking about. He's also been one of the loudest voices within the so-called singularity community, and has gone so far as to chart out a graph of real-world technological growth and compare it to what Moore's Laws predicted. Remember how I said Moore's Law outlined the advancements made in the past few decades, and how it lined up shockingly well with tech that's now being released? Well, Kurzweil noticed this back in the 80s and charted out how technology would continue to improve for decades to come. Using these numbers, Kurzweil predicted and maintains, that the singularity will occur around 2045, thirty years from now [2015]. So going back to our original question here, is Terminator's future possible where Skynet takes over in 2017? Oh, that dates off the mark, just like its predecessors were before it. Skynet’s not coming in two years, it's coming in 30.

Trivia

 * This episode was sponsored by Audible.