How realistic are famous movie robots?

Mark Lankester
Yahoo UK Movies Features
Iron men... classic movie robots (Credit: Rex Features)

Robot and Frank’, out on DVD 15 July, gives us perhaps the most realistic movie robot so far. Not only does it look like robots that actually exist right now, it also has a plausible job – looking after an elderly chap (played by Frank Langella).

[Arnold Schwarzenegger set to return as The Terminator]

Many movie robots have been a bit more fanciful however, according to renowned furturologist Ian Pearson. Pearson helped develop text messaging for BT back in the 90s, and predicted the likes of search engines and digital TV. So, we enlisted his sage-like powers to find out if some of the robot scenarios predicted by Hollywood – such as Terminators wiping us all out - could ever happen in reality. (For our sake, let's hope that one's a "no"...)

Robot from ‘Robot & Frank’
Operating history: Ageing ex-convict, Frank, is bought a robot pal by his family, which helps him with therapeutic care and his daily routine. However when Frank finds out the robot is not programmed to distinguish between legal and illegal activities, he uses his new companion to help him commit one last heist. But with Frank needing to cover his tracks, he’s faced with wiping the memory of his now respected robotic friend, as it could be used as evidence against him.
Could it happen? “Robot’s design is actually similar to Honda’s ASIMO robots from Japan, already grounding it in reality," says Ian. "We already have basic robots doing essential caring jobs, and they’re capable of simple conversation. Even some basic A.I. systems are capable of almost fooling us into believing they’re people (think online Chatbots). Although it might seem expensive to use a robot where a human might do, in reality it’s not about economics, it’s about a shortage of people. In Japan for example, where these kinds of robots are being developed, they have a lot of older people coming along, and not enough younger people to care for them. I would say the realistic timeframe for this to start becoming widespread is between 2020 and 2025. It’s not very far away.”
Robo-realism rating: 5/5

Johnny 5 from ‘Short Circuit’
Operating history: Originally proposed by the military for handling nuclear weapons, prototype Number 5 is struck by lightning - developing consciousness and a fear of death (reprogramming) – causing him to flee his creators.
Could it happen? “We already have bomb disposal robots,” says Ian. “Why would you take one of those a stick a head on it? Plus, if you’ve got the level of technology to create complex A.I. with conversational abilities, you wouldn’t see the rest of the robot based on caterpillar tracks. It’s just not feasible.”
Robo-realism rating: 2/5

The Nexus-6 Replicants from ‘Blade Runner’
Operating history: Genetically engineered organic robots, known as Replicants, are almost indistinguishable from humans by the year 2019. They are used for dangerous or leisure work in the Off-World Colonies, but banned back on Earth. Then, when a group of particularly ruthless Replicants escape into LA, a retired detective (or Blade Runner) is brought in to hunt them down.
Could it happen? “There’s currently a lot of work going into trying to create robots that look like humans, however there’s the issue of ‘The Uncanny Valley’," says Ian. "When you’ve got a robot that looks quite lifelike, but not exactly right, then it becomes a bit ghoulish and very obviously not human. People tend to react with distaste and even disgust (think dodgy wax-works). However the ones in ‘Blade Runner’ have overcome that problem and are now so human it’s become worrying. The problem with ‘Blade Runner’ though is they dated it too early (2019), in reality we could have this kind of robot routinely running around by about 2070. What the movie did get right was the excessive strength, it’s entirely realistic. Weight for weight the technology would allow them to be about five times stronger, making them prime for use in areas such as the military and emergency services.”
Robo-realism rating: 3/5

NS-5, a.k.a Sonny from ‘I, Robot’
Operating history: In 2035, robots co-exist with mankind as personal assistants and public service workers, installed with ‘Three Laws’ to prevent them from harming humans. However when the founder of manufacturers US Robotics is killed in an apparent suicide, prototype bot Sonny, who is able to override the ‘Three Laws’, comes under investigation.
Could it happen? “These robots are very feasible indeed," says Ian. "You’ve got basic humanoid robots, and they’re not trying too hard to make them look exactly like humans, although they’ve do have internal projected faces so they avoid the ‘Uncanny Valley’ problem. However you can image these robots being quite commonplace in the not very distant future, then towards the later part of the century they would be replaced by ‘Blade Runner’ style robots, once the technology caught up.”
Robo-realism rating: 4/5

David from ‘A.I. Artificial Intelligence’
Operating history: In the late 21st century, advanced humanoid androids known as Mecha are capable of mimicking thoughts, emotions and actions - making them suitable for use as toys, companions and even lovers. Prototype child mecha (David) is produced, with the ability to show genuine and unconditional love, but the experiment quickly backfires.
Could it happen? “David is cute as a character, especially the ‘Pinocchio’ complex he acquires, but the way the movie tackles emotions belittles the way we would actually deal with A.I.,” says Ian. “However Teddy, that wonders round as a companion for kids, is perfectly realistic. It’s not too dissimilar to Furbies as a toy concept. Also, the sex-bot, Jude Law’s adult gigolo robot, is perfectly feasible too - and a likely eventuality in our future. Of course robots will develop emotions, as in reality an emotion is just an imbalance in the way we make a decision. It’s just a modifier in an equation, a virus if you like. There are many different ways robots could be programmed to mimic these - in fact there’s nothing in principle that would prevent a robot existing with the capabilities to show millions of different types of emotion. Far more than humans themselves. Interestingly, just as we’ve just been through the battles over same sex marriage, how long will it be before someone wants to marry a robot? That will happen. It’s just a question of when.”
Robo-realism rating: 4/5

[Prometheus sequel to feature robot-aliens?]

The Terminator, T-800 Series from ‘The Terminator’
Operating history: Military A.I. network Skynet becomes self-aware in 1997, defending itself by launching an all-out nuclear attack on the perceived threat that is humanity. By 2029 mankind is fighting for survival, as “The Machines” send a ruthless robotic assassin, The Terminator, back in time to 1984 to kill Sarah Connor - would be mother of John Connor, the leader of the human resistance.
Could it happen? “The IT industry actually started worrying about the ‘Terminator Scenario’ of self-aware machines about 10 years ago,” says Ian. “But we came to the conclusion that there’s more to intelligence than number crunching  -so digital machines will never have a chance. It’s more likely with analogue machines as there’s the potential to function more like a human brain. However, it’s the idea of ‘killing machines’ that we have to really worry about. Of course we already have these in the controversial form of military Drones, but all that’s missing is the A.I. for them to operate independently. It could be done, in the next 5-10 years, but of course nobody in the military actually wants to. The ‘Terminator Scenario’ is however possible as an accident: say a group of A.I and Biotech students attempting to create a self-aware system as an experiment, and then it just goes out of control. Creating a self-aware machine is a genuine goal for some developers, but it won’t become a realistic scenario until at least 2025 - it's around then we will have to start worrying about the capabilities and consequences of doing so.”
Robo-realism rating: 4/5

Wall-E (Waste Allocation Load Lifter – Earth Class) from ‘Wall-E’
Operating history: In the distant future, mankind has abandoned an over-polluted Earth, leaving a group of waste-management robots (Wall-E units) to clean up the mess, ready for humanity’s return. By 2085 just one Wall-E remains, having developed sentience after 700 years of existence.
Could it happen? “A rubbish-compactor robot that runs on sunlight? Wall-E just isn’t feasible,” Ian says. “In reality, to deal with that kind of a job, Wall-E would actually be a whopping great big truck. Sunlight just wouldn’t provide the power levels to do that work, especially in a post-apocalyptic world where sunlight might not even exist. Why not run it on the rubbish?”
Robo-realism rating: 1/5

Robocop (formerly police officer Alex Murphy) from ‘Robocop’
Operating history: In the near future, Detroit is in dystopian ruin, overrun with crime. However when veteran cop Murphy is brutally gunned down by a criminal gang, corporate giants OCP harvest his body to create a solution to Detroit’s problems – the first ever police cyborg, Robocop. But how much of Murphy’s past identity remains?
Could it happen? “Things become confusing once you start deal with cyborgs (part human part robot), but there are arguments for retaining the human element,” says Ian. “There are already autonomous robotic guards being used in some policing systems around the world, and although they can fire non-lethal weapons, it’s been pointed out that they could easily be adapted as battlefield robots. But whereas on a battlefield the objective is usually fairly obvious, policing, in theory, is far more sophisticated; involving detection, understanding, judgment and even discretion. A robot could easily detect an intruder, but when it comes to simple crimes such as littering, where something may or may not be an accident, it becomes far more complex. I wouldn’t say we won’t ever have policing robots, but give it maybe 40 years and we’ll be closer. We’d need to have computers educated in human culture before they can understand the issues involved in judging it – then we could have the Robocop argument.”
Robo-realism rating: 2/5

Data from ‘Star Trek: Generations’
Operating history: A fully sentient android who serves as an officer on the USS Enterprise. In ‘Generations’ he activates an ‘emotion chip’, which he initially struggles to control.
Could it happen? “He’s supposed to be unique and he’s supposed to be mid-24th century," says Ian. "But in one episode of 'Star Trek', Data is tried by a human rights court to decide if he has to right to refuse to be used for experiments. Data's then asked to list his performance specification, but the processor and memory capability he mentions, our computers passed in 2005! The writers don’t really know much about what is feasible with A.I. By the time we reach ‘Star Trek’, society will have passed robots of that complexity, or better, for hundreds of hears before then - in reality most of the crew of the Enterprise would be robots. Also, if you’ve got a robot with that sort of capability, why would you handicap it by giving it a monochrome face? It looks wrong. Because he’s so superior to everybody else in the crew, why is he just one of the staff? When you think about it, why is he not the captain?”
Robo-realism rating: 1/5

'Robot & Frank' is out on Blu-ray and DVD July 15.