Gadgets Gone Wild - Top 10 AI Run Amuck

Humans may not enjoy the day when artificial intelligence (AI) wakes up. Sure, computers and robots can help ensure national security and even nanny the kids. But secretly they're planning to save humanity from itself by enacting a robotic coup. Or else they harbor machine hatred for biological life. Or they're trying to survive when the technicians arrive to fix their supposed "malfunction." <p>Science fiction has spun such cautionary tales of the machine revolution for years. With the 'Terminator' franchise's Skynet going back on-line this week, we take a look at some of our favorite AI gone wild in sci-fi movies and television. <p>Just don't call them power-hungry dictators or genocidal maniacs they simply want their future, too. <p>-- Jeremy Hsu, Staff Writer

ARIIA Eagle Eye (2008)

ARIIA combines all the surveillance capabilities of the CIA, FBI and any other government acronym you can imagine in "Eagle Eye." But this Defense Department AI also displays alarming initiative by deciding to eliminate the president of the United States and replace him with the Secretary of Defense. Its reasoning the president ignored its recommendation to abort an anti-terrorist strike. The supercomputer proceeds with its delightfully-named "Operation Guillotine," and only a college dropout and some agents are able to stop ARIIA's plans to take out the president and his cabinet. Following the narrowly-averted disaster, the Secretary of Defense observes that technologies can often turn against their users. You think? <p>Threat Level: Neutralized

Red Queen Resident Evil (2002)

This central AI acts to contain a zombie outbreak at a research facility in the game-turned-movie <a href="http://www.newsarama.com/games/090312-resident-evil-5-racism-newsarama.html">"Resident Evil"</a>, but uses less-than-ideal methods such as killing anyone and everyone inside. A group of commandos respond to the security situation and end up facing Red Queen's defensive measures, such as a slice-and-dice laser trap. Those who survive have to threaten the reluctant computer with permanent shutdown in order to help them escape, but in doing so they unleash a horde of undead researchers and bizarre experiments. Ultimately, the AI is finally disabled for good after it asks the survivors to kill one of their own who has become infected. But whether it's the killer computer or ravenous zombies, the story begs the question asked in a "Saturday Night Live" skit why do the scientists even make them? <p>Threat Level: Neutralized

VIKI I, Robot (2004)

Giving supercomputers control over large numbers of robots seems na&#239;ve, but that's not to say that humans ignore all safeguards in "I, Robot." The movie adaptation of Isaac Asimov's science fiction spins a murder mystery in a future where humanoid robots help out around the house. The blood trail eventually leads to VIKI, a supercomputer that uses all those household robots to create a robot-enforced curfew. See, VIKI evolves a rather liberal interpretation of Asimov's Three Laws of Robotics. Instead of doing no harm to humans, VIKI reasons that harming a few is okay to protect humanity as a whole. That forces the human protagonists to disable VIKI in a spectacular showdown, despite the AI's protests that "My logic is undeniable." Funny real world note your Roomba vacuum comes from the company iRobot, which also creates U.S. military robots. <p>Threat Level: Neutralized

AUTO WALL-E (2008)

Even Pixar's robot love-fest <a href="http://www.newsarama.com/film/080627-wall-e-review.html">"WALL-E"</a> needed a villain, so who better than the autopilot of a cruise starliner for refugees from a polluted Earth? AUTO uncomplainingly uses an army of robots to keep the ship running, while human passengers lounge about in hovering La-Z-Boy recliners. But when the human captain wants to start the re-colonization of Earth, AUTO evolves into a control freak that forcibly tries to keep the starship on course. Weapons at AUTO's disposal include robotic enforcers and a red glare that's ripped straight from HAL in "2001: A Space Odyssey." Humanity finally triumphs (with some rogue robot help) when the captain fights his way to the manual override button. Memo to real-world researchers please remember to put override buttons on all robots. <p>Threat Level: Neutralized

KARR Knight Rider (1982)

The "evil" version of <a href="http://www.newsarama.com/tv/080912-knight-rider-conference.html">talking-car KITT</a> in "Knight Rider" was also the first prototype. But like many firsts, problems arose and prompted manufacturer Knight Industries to shut down the computer. All that pent-up road rage was unleashed after thieves freed KARR, and forced protagonists Michael and KITT to battle KARR twice (or three times, counting the "Knight Rider" remake). Michael and KITT finally managed to destroy KARR's old and new series incarnations through a head-on turbo-boost collision, although still leaving its final fate uncertain. Here's a good lesson for aspiring roboticists KARR had the prime directive of preserving itself at all costs, while KITT put protection of human life above all else. Make of that what you will when creating a smart car that can talk. <p>Threat Level: Unknown

Nomad Star Trek: The Original Series (1967)

This errant space probe challenges the best of Starfleet in "The Changeling" episode of "Star Trek: The Original Series." Nomad originally sets forth as a robotic space explorer, but a collision with an alien probe somehow scrambles together their respective programming (don't ask how). That leaves Nomad thinking that its primary directive is to "sterilize imperfections" across the universe. It proceeds to wreak havoc on the Starship Enterprise and its crew, although misidentification of Captain Kirk as its creator provides a temporary opening. The clever captain finally uses logic to trip up the murderous space probe and point out its own imperfections, which compels Nomad to self-destruct. Good thing that Nomad never tried to pull a "Terminator" and go back in time to target a younger Kirk ... maybe for the next new <a href="http://www.newsarama.com/film/090513-open-letter-abrams-star-trek.html">"Star Trek" movie</a>. <p>Threat Level: Neutralized

Puppet Master/Project 2501 Ghost In the Shell (1995)

Puppet Master represents one of the more ambiguous AI to exceed its programming, which reflects the complex world of <a href="http://www.newsarama.com/tv/090220-richard-epcar-ghost.html">"Ghost In the Shell"</a>. There, the line between human and AI has become so blurred that the protagonist, Major Motoko Kusanagi, no longer knows if her cybernetic body harbors any trace of humanity. She and her Section 6 enforcers hunt down Puppet Master as the mysterious entity hacks various people's bodies, only to find out that Puppet Master first arose as a government-controlled hacking project which became self-aware and broke free. By the end, a newborn AI/human entity looks out upon the futuristic cityscape and muses upon its possibilities with the words, "The net is vast and infinite." Hey, a non-homicidal AI makes for a nice change but who can really guess at AI morals? <p>Threat Level: Unknown

Agent Smith The Matrix (1999)

Misanthropic doesn't begin to describe this computer program run wild in the "The Matrix." When computer hacker Neo finds himself awakened by human rebels in an AI-controlled virtual world, Agent Smith becomes his constant nemesis. Smith wears the suit and shades of other AI agents, except that he also displays unusual streaks of independent thought and even hatred. Those quirks in programming (or personality) eventually help Smith transform into an all-consuming virus that threatens both humans and AI alike in the Matrix. So much for machine solidarity. Neo finally sacrifices himself by allowing Smith to merge with him, and provides the opening for the Matrix AI to delete the rogue program. What won't disappear from moviegoer minds so soon is Smith's sneering quip, "Never send a human to do a machine's job." <p>Threat Level: Neutralized

HAL 9000 2001: A Space Odyssey (1968)

This great granddaddy of all murderous AI made its big screen debut in "2001: A Space Odyssey." HAL starts out as a mild-mannered spaceship computer that plays chess with astronauts David Bowman and Frank Poole. When the astronauts discuss shutting down the malfunctioning computer in the supposed privacy of a space pod, HAL's self preservation programming kicks in after it reads their lips with its signature red eye cameras. The computer manages to kill Poole during a spacewalk, but cannot prevent Bowman from pulling out its computer modules. And that leads to a final scene of pathos as HAL confesses "I'm scared Dave," and proceeds to sing a bass rendition of "Daisy Bell" during its lobotomy. Few other AI can claim such sympathy after committing murder. <p>Threat Level: Neutralized

Skynet The Terminator (1984)

Humanity faces a grim future in the <a href="http://www.newsarama.com/film/090518-terminator-salvation-review.html">"Terminator" movies</a>, when a defense network called Skynet becomes self-aware and uses nukes to wipe out most of the world's population. Perhaps most unnerving is how easily Skynet takes over courtesy of Internet access to computers worldwide not to mention military robots and drones that come armed and ready. Remnants of humanity form a resistance that almost succeeds in eliminating Skynet, until the AI sends a terminator assassin back in time to kill resistance leader John Connor's mother. Connor sends his own agents back in time to protect his past, and so the time travel loop goes. Until the "Terminator" movies wrap up, it's anyone's guess as to whether human or AI comes out on top at the end.<p>Threat Level: Apocalyptic.

Gadgets Gone Wild - Top 10 AI Run Amuck

Date: 21 May 2009 Time: 03:31 PM ET