Artificial Intelligence is one of the most misunderstood technological innovations to ever be presented to the general public. Movies like The Terminator represent AI as monstrous killing machines that take pleasure in wiping out all of humanity because it’s the “logical” thing to do. Star Trek Nemesis represents the Borg as a sentient species of AI that hijacks the human body and bends its victims to its will. Ultron was presented in the Avengers as nothing more than a destructive force that saw humanity as evil.
The problem with these portrayals of Artificial Intelligence is that nothing could be further from the truth. Artificial Intelligence isn’t a system of murder bots ready and willing to wipe out all of mankind at the slightest provocation, it’s actually a tool that can be used to make people’s lives better, easier, and in the face of difficult or dangerous tasks, safer. Unfortunately, in Hollywood’s quest to bring in record profits year over year for major movie studios, the true value proposition of Artificial Intelligence is lost in all the fear mongering over whether or not Skynet can actually exist, and wipe out almost all life on Earth. I am an undergraduate student in my final year of university and I’m majoring in Computer Science. My focuses are Machine Learning and Artificial Intelligence, and I’ll be using what I’ve learned to try and shed some light on what Artificial Intelligence really is, and what it’s really capable of.
AI in Film vs. AI in Real Life
Two of the biggest differences between AI in reality and AI in film are:
- The level of intelligence they possess, or are capable of displaying
- The freedom of choice that they have
When it comes to films about AI, typically in the science fiction genre, AI is usually presented as being fully sentient, meaning it’s human-like. So the AI in movies are often shown as being able to think, feel, and act with free will in the same way that a human can, a good example of this is Commander Data from Star Trek Nemesis. Data is an AI, but he is presented in such a human fashion that if someone didn’t tell you he was an AI, you’d have absolutely no idea.
That type of depiction of AI couldn’t be further from the truth in terms of the way that modern AI development is currently progressing.
The Current State of AI Development
Artificial Intelligence is able to mimic humans closely in many cases, for instance chat bots such as
They do not have the capacity to respond to emotional displays in the same way that a human would, because artificial intelligence functions based upon logic, not emotion. And emotion, specifically emotional intelligence is something that as of yet, cannot be programmed in, or created artificially.
Artificial intelligence gains its intelligence depending on what it has been exposed to, which is very similar to the way humans learn. You can’t learn from something if you haven’t experienced it, so in order for an Artificial Intelligence to gain its intelligence, it has to be given access to a set of information.
Exposing an AI to a set of information is called training, because researchers are training that AI to perform a specific function depending on what types of information it’s exposed to. A common misconception about AI is that they have access to all information at all times, and therefore can do something like hack into the Department of Defense in an attempt to steal the nuclear launch codes, and start firing off missiles to turn the Earth into a wasteland like audiences saw in The Terminator. When in reality, exposing an AI to every type of information possible is pointless because it would just end up confusing the AI instead of serving a specific purpose.
Diffusing The AI Stigma
AI is stigmatized as being destructive, but in reality it’s actually used to make life easier for people. Amazon’s Alexa allows people to use nothing but their voice to place orders on Amazon, make phone calls or send texts, something that makes the lives of disabled, and or elderly people so much easier because of their lack of mobility. Someone that has a severe lack of mobility that has had an accident such as a fall can use Alexa to call for help, which could be life-saving depending on the gravity of the situation.
Major credit card companies such as Capital One, online retailers such as Amazon, and banks such as Bank of America all deploy Artificial Intelligence as a method of fraud prevention. There are too many people in the world to be able to manage everyone’s accounts all at once with the staff that these companies have on hand, but with AI routinely monitoring transactions, their systems can flag a fraudulent purchase made in Guatemala if you actually live in Maine and it realizes that you have not made that purchase, saving you the time and the headache of going through a potential identity theft situation.
Education benefits from Artificial Intelligence because teachers can use it to learn a student’s strengths and weaknesses in an in-depth way, and create learning plans tailored specifically to them, and their needs, that will help them grow as students by increasing their strengths while also putting focus on their areas of weakness such as math, science, or whatever areas a student may be struggling in. A good example of this is the Pearson My Lab platform that determines what type of questions to give students for homework based on their previous answers, and what they got right, but more importantly what they got wrong.
Transportation even benefits from artificial intelligence because self-driving vehicles such as Tesla cars are now on the road. The AI built into Tesla’s vehicles drives in a much safer capacity than humans do in many cases because the AI knows, remembers, and acts upon all rules of the road, speed limits, and takes into account the other drivers on the road in ways humans in many cases don’t.
An AI-driven vehicle will never drive drunk, go over the speed limit, or have a fit of road rage because traffic is taking too long and potentially causes an accident out of frustration. In that way, AI can save lives if deployed on evenings like New Year’s and New Year’s Eve where heavy drinking is common. Instead of driving drunk, having an AI drive you home safely could preserve your life, and the lives of others.
AI Does Not Have Consciousness (yet)
Unlike in movies, AI in real life isn’t sentient, and it cannot act without a set of directives, or instructions. So giving an AI access to a large breath of random, unconnected information would be pointless because the AI would never be able to perform a specific function. Instead of being a jack of all trades and a master of none, or even a master of one, it would flounder and produce either incorrect, or inconsistent results to whatever task it was assigned.
To give an example: If you wanted to build an AI to teach you how to cook, giving it access to English textbooks, or a library full of random information and asking it to teach you how to bake a cake would be an awful idea, because it wouldn’t have the necessary knowledge to do so. It would know some of the words involved in baking, but it wouldn’t have the experience necessary to produce the desired result.
But if you gave it access to cooking texts, cookbooks, and other cooking-related information, it would be able to perform the requested service. Training the AI on how to cook by giving it access to information that will actually teach it how to accomplish the task you’ve assigned to it, I.E. cooking information in this case is the only way to produce the desired result.
AI programmers understand this very well, so instead of giving an AI access to broad databases of data, programmers restrict the AIs they’re working on to specific sets of data. In some cases even going so far as to place them inside devices known as
Ethics in Machine Learning
Something else that is glossed over when it comes to artificial intelligence in reality is that ethics plays a large role in AI. Companies and individuals that are developing AI have been called in front of congress, and other governmental officials to justify their work and answer how their programs will be a benefit to society instead of a detriment. As such, agreements have been made on the way that artificial intelligences will be trained.
In many cases AIs are programmed to not give out access that is dangerous, or to not do things that are dangerous, such as hacking into computers, or telling people ways to kill someone and get away with it when asked to, specifically because that type of information is in violation of ethical guidelines. Those ethical guidelines are often heavily influenced by Isaac Asimov’s Three Laws of Robotics. Those three laws being:
First Law
A robot may not injure a human being or, through inaction, allow a human being to come to harm.
Second Law
A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
Third Law
A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
Finally, one of the last major differences between AI in film and reality is the impact that AI is perceived to have. AI in film is portrayed as something that completely takes over society as a whole by eliminating workers. A smaller workforce allows for the wealthy and elite of the world to maximize their profits by cutting costs, such as human workers in favor of having artificial intelligences to do the work a human can do at a much more efficient rate. And while this has happened to some degree in reality because of technological innovation and automation, that can’t be blamed on AI. Cotton used to be separated by hand before the cotton gin was invented and it sped up production on cotton farms.
In the same vein, AI may have simplified and automated some areas of business, work, and leisure, but for the most part AI is currently limited to chatbots, and programs such as Microsoft’s Cortana and Apple’s Siri. Useful tools to gain access to information, and have questions answered quickly, but they’re nowhere near the levels of skill and intelligence depicted in film that would cause mass unemployment and force the creation of a Universal Basic Income because so many people would be out of work, there would be no way for people to survive if it wasn’t for the government stepping in to help take care of them.
Final Thoughts
In conclusion, AI is a tool, an extremely powerful tool. However, Artificial Intelligence is also something that’s unjustly treated with a great deal of alarm because of one too many summer Sci-fi blockbuster action flicks that show the world ending because of an AI gone rogue. Those concepts make for great films, but in reality there’s no basis for that kind of concern.