IoT ruins movies: The Terminator
Setting the scene: Artificial intelligence and robots
At the heart of the movie The Terminator is a fundamental struggle between man and machine. In 2029, the future from which the Terminator is sent back to wreak havoc on the life of Sarah Connor, a state of perpetual and horrifying warfare exists between human beings and the machines they trusted to protect them.
The machines, we are told, got ‘smart’. They are hooked into everything. Self-aware. And they’ve decided they can do very well without humans, thank you very much.
OK, one or two uncomfortable similarities between this state of affairs and the IoT did make me squirm a bit, if only because the terminology is so alike. ‘Smart’ objects. ‘Intelligent systems’. ‘Machine learning.’ Eek.
But before we give humanity up for lost and scamper off to our bunkers, I thought it might be fun to see just how far these similarities go, by comparing the Terminator to some of the humanoid bots already in use. Luckily, there’s one crucial difference between the Terminator and the IoT bots that will save us all – but you’ll have to read the rest of the blog to find out what.
The movie: The Terminator
Directed by James Cameron and starring Arnold Schwarzenegger and Linda Hamilton, The Terminator is a 1984 science fiction action film that topped the US box office for two weeks and quickly passed into the public consciousness as one of the most significant sci-fi films to date.
Schwarzenegger plays the Terminator, a cyborg assassin sent back from 2029 to kill Sarah Connor, whose son John will eventually lead a resistance against the machines that have exterminated mankind. Sarah has a fortunate ally in Kyle Reese, a soldier from the future send back to protect her.
Humanoid robots: Pepper and Connie
While nothing as sophisticated as an indistinguishably human-looking robot exists yet, the world of IoT has given us some smart bots already. Pepper, co-developed by Aldebaran and Soft Bank Robotics, is an emotional humanoid bot – so called because it is designed to recognize human emotions through tone of voice and context.
Another is Connie: a concierge robot also from Aldebaran and Watson, which is currently being tested in some Hilton hotels. Connie knows everything there is to know about the hotel environment, and can answer guests’ questions and make recommendations.
The Terminator: Humanity is only skin deep
While Pepper and Connie are built to augment and aid the human, the Terminator’s human attributes are only skin deep. Flesh and blood cover a machine endoskeleton, but there the humanity ends. The Terminator feels neither pity nor pain. It cannot recognize or respond to human emotion of any kind, because it doesn’t need to.
In contrast, Pepper and Connie are designed to do exactly that. Both are equipped with Watson Tone Analyzer, which uses linguistic analysis to detect three types of tone in written text: emotions, social tendencies, and writing style. Tone Analyzer can understand emotional context of conversations and communications and use this insight to respond appropriately. Since emotional intelligence is a fundamentally human advantage and would be of no use to the bots themselves, it follows that the fundamental purpose of these bots is to serve humans. In other words, Pepper and Connie are designed for humans, by humans.
What are the similarities?
There are some similarities to the IoT bots, however, and we can spot some existing capabilities in the Terminator’s technical arsenal:
- Sensors: The Terminator uses a sophisticated array of sensors and machine learning capabilities to make sense of its immediate environment.
- Speech to Text: Like Pepper and Connie, the Terminator is able to take large, unstructured data sets, like overheard conversation, and translate it to text that can be analyzed.
- Natural language processing: Natural language processing allows the Terminator to ‘understand’ language by comparing the likelihood of various alternative meanings within a given context – something that IoT bots are also able to do. Background data enables them to make sense of context in order to determine a word’s most likely meaning: whether ‘knife’ refers to a weapon or a piece of cutlery, for instance.
- Overlaid real-time information: last year, the US cycling team were trialing Solos eyewear – glasses designed to display real-time data on-screen, while enabling the wearer to still see where they were going. The overlaid data stream relays information about the cyclist’s physical condition (heart rate, and so forth) and progress (km travelled, average speed, incline) in real-time. The Terminator employs something similar – his infrared vision is overlaid with what looks like computer script – feeding back information about his own processes as they take place.
- Machine learning: Machine learning refers to the ability for ‘things’ to learn and adapt as their acquisition of data grows. The Terminator demonstrates many instances of this skill, most memorably when he picks up the expression “F*ck you, asshole” from an early assailant and deploys it (to comic effect) later in the film.
The crucial difference: trust and decision making
Luckily for IoT fans (and the human race in general) there is one, crucial difference between IoT bots like Pepper, and the Terminator. Pepper and Connie are not trusted with decisions. They can make recommendations based on their programmed data and the real-time information they are able to acquire, but the decision to act always rests with a human being. The Terminator, on the other hand, can take decisions for himself, putting him in the unprecedented position of being both computer and programmer. He is answerable to no-one but himself, and therein lies the danger. If the Terminator had been an IoT bot, he could never have started his rampage against Sarah Connor – at least not without a human’s say-so. Comforting thought, that.