World – In the year 1950, John McCarthey, computer scientist, invented the term Artificial Intelligence (AI). He believed every aspect of learning or other features of intelligence could be so precisely described that a machine could be made to do it. As we find out by looking at our lives, McCarthey was seemingly right. AI is one of the emerging technologies. While there is a chance it can become exactly what McCarthey dreamed of, AI as it stands, is still some way off from that dream.
Before we delve into how AI is helping humans to advance, one must first understand what it is. In simple terms, AI is the ability of a computer program to learn and think. The ability for it or a robot to perform tasks commonly associated with intelligent beings.
The different forms of AI
There are three main types of AI based on its capabilities. Weak AI, also known as narrow AI, strong AI, also known as general AI and super AI. While many believe AI is just highly advanced robots that could turn on us and destroy civilization as we know it, that is not the case. We actually use it more often than you think
Weak AI describes the algorithms that can complete a set of finite functions. Although it can solve complex problems and perform them more efficiently than humans, its functionality is severely limited to its programming. It can only be programmed to do one task and one task only. The idea behind weak AI isn’t to mimic or replicate human intelligence. Rather, it’s to simulate human behaviour. We use this AI every day when we use Siri or Alexa, search for shows on Netflix or use Google. Currently this is the only AI that exists to this day. Although some might view it as barely intelligent, it’s highly intelligent at completing the specific tasks it’s programmed to do.
Then we have strong or general AI. Strong AI refers to AI that exhibits human-level intelligence. This AI can understand and learn any intellectual task a human being can do. It should also be able to reason, solve puzzles and exhibit common sense among others. In theory, anything a human can do, strong AI can do. It is theorised that for researchers to reach strong AI, machines would need to be made conscious.
The last is super AI which is still a concept. Super AI will surpass human intelligence and ability. It’s also known as artificial superintelligence (ASI) or superintelligence. Even the brightest minds cannot come close to it. An example of this would most likely be Terminator from Terminator or, if you are into video games, the AI in the Mass Effect trilogy would also be a good example of it.
AI based on functionalities
To describe various types of AI, it is necessary to categorise them based on their functions. The most basic variety of AI is called a reactive machine. It merely reacts to current scenarios and cannot rely upon taught or recalled data to make decisions in the present. It does not plan anything in advance or based on previous information and merely watches the current environment and acts on it. They are given specific tasks and don’t have capabilities beyond that. Deep Blue, IBM’s chess-playing supercomputer, is the perfect example of this type of machine. It watches the moves of a player and then predicts what moves they will make next and chooses the optimal move to play and doesn’t have any concept of the past, nor any memory of what has happened before. It perceives the world directly and acts on what it sees at that moment.
Then there is limited memory which can look into the past. Self-driving cars use this technology to observe other cars’ speed and direction. Unlike reactive machines, this can’t be done in one moment. Cars and specific objects are monitored over a period. These observations are added to the self-driving cars’ preprogrammed representations of the world, which also include lane markings, traffic lights and other important elements, like curves in the road. This information helps to inform a car when it is safe to change lanes to avoid cutting off a car, when to turn and how to turn. Unlike humans, this information disappears. No amount of driving will make the car build up ‘experience’ as humans do.
Theory of mind AI is the next step in AI evolution. In psychology, the Theory of Mind is the ability to understand other people and their behaviour by attributing mental states to them. This is the divide between the machines of today and the machines of the future. The hope is machines of the future will have this understanding as well. Such AI requires a remarkable amount of understanding of humans, things and the environment. The Kismet robot head, developed by Professor Cynthia Breazeal, could recognize emotional signals on human faces and replicate those emotions on its own face. Humanoid robot Sophia, developed by Hanson Robotics in Hong Kong, can recognize faces and respond to interactions with her own facial expressions. These are very rudimentary examples. They are still extremely far off from truly understanding human emotions, needs, wants, beliefs and behaviours.
The last is self-awareness AI which only exists hypothetically, which some are very thankful about. This will require researchers to not only understand consciousness but build machines that have it. It is an extension of the theory of mind AI. consciousness and self-awareness is the same thing. Conscious beings are aware of themselves, know about their internal states, and are able to predict feelings of others. For eg. We assume someone shouting means they are either angry or displeased because that’s how we would feel when we shout. Without having a theory of mind, we would not be able to come to this conclusion. Self-awareness AI will be robots who are exactly like humans but far superior. They will be able to understand human emotions, needs, wants, beliefs and behaviours and have them as well. They might not even believe they are robots but humans. The perfect way to describe it is by looking at the ‘Synths’ in the video game Fallout 4. To the naked eye, they are human beings in every way but underneath the flesh, they are just mechanical beings.
Reducing human error
The term human error was created because humans make mistakes, sometimes costly ones. Computers, if programmed correctly, don’t make mistakes. The decisions taken by AI are based on information previously gathered and a certain set of algorithms. Overtime, as more data is gained, errors can be reduced to zero and the chance of reaching accuracy with a greater degree of precision is a possibility. People also get tired which leads to costly errors, machines do not. This also eliminates any errors that can occur from fatigue.
Zero risks for humans
One of the biggest advantages of AI is that humans can essentially forget about putting their lives at risk. Why would anyone need to defuse a bomb or rush into a nuclear reactor or travel to deep space when there is a robot to do it for you? AI can be used effectively to handle disasters like forest fires before they get out of control.
For too long have we strayed from exploring the darkest depths of the oceans and the shining galaxies just beyond our reach because of the dangers associated with it. With AI, these dangers no longer exist. Man can venture forth and explore the unknown.
Always available to work
Humans can only work effectively for a couple of hours before needing a rest or they run the risk of burning out. AI have no need to rest and can be made to work endlessly. Running 24/7, 365 days a year. Educational institutes and helplines can effectively be open 24/7 with the help of AI. They can even think much faster than humans and perform multiple tasks at a time with accurate results.
Deeper data analysis
The modern business and even sporting worlds are swimming in data. This data is crucial to any company or sports team but how the data is analysed is even more crucial. Manual data analysis is a time-consuming enterprise but with AI, it can be processed at an incredible rate. AI systems can analyse the data and provide valuable feedback which can be crucial to marketing, identifying customer preferences and much more.
Unbiased, informed and faster decision making
Humanity’s greatest strength is their emotions, but their greatest weakness is also their emotions. When we make a decision we analyse many factors both emotionally and practically. This not only slows us down but can cloud our judgement. An AI-powered machine takes in only the data provided to it to come to an unbiased, informed and due to its processing power, a faster decision. Emotions play no part in its decision and therefore its choice will purely be based on what is best.
New inventions that use AI
In practically every field, especially the medical field, AI is the driving force that is being used to create a number of inventions that will aid humans. The current medical industry has used artificial intelligence in a significant way in Radiosurgery. The AI helps to operate on tumours without damaging the unaffected surrounding tissues.
Then there is Trailjectory which uses AI to empower patients on their cancer journey. It analyses its global patient community’s accumulated data so cancer patients can make informed decisions. It analyses all relevant treatment options and instantly presents what is relevant to the concerned patient.
Disadvantages of AI
At the start I mentioned AI has not yet been perfected. One of the biggest disadvantages of AI is the extremely high cost to create it and fund research on it. AI also needs to operate on the latest hardware and software to stay updated and meet the latest requirements, thus making it quite costly. Being a relatively new technology and constantly evolving, there are huge costs to it.
The costs of jobs
AI is great to do repetitive jobs but you know what else is good at it? Humans. We have already seen how the automobile, canning and other industries have laid off workers because it is cheaper to have machines do the jobs. No mistakes and no need for breaks make machines the perfect tool to cut costs and maximise profits. As AI progresses, more and more skilled workers could see themselves losing their jobs.
No ethics and emotionless
Since we haven’t reached self-awareness AI, humans are the only ones capable of having ethics, morality and showing true emotions. This makes us unique. The connections we build between each other help to create communities. There is a growing concern AI is progressing too quickly and could lead to the extinction of the human race. This even has already been termed as the AI singularity. The moment AI becomes so self-aware, it decides to launch an attack of those who created it.
When AI malfunctions
AI is still like any piece of technology you have around your home. It can be hacked, disrupted or even malfunction. Some malfunctions are more costly than others. On 18 March 2008 Elaine Herzburg became the victim of the first recorded pedestrian fatality involving a self-driving car. Herzburg was fatally struck by the Uber test vehicle whilst pushing a bicycle across a four-lane road. After this, Uber suspended the testing of self-driving vehicles in Arizona where the incident took place. This is not an isolated incident. A study by the University of Michigan found that autonomous vehicles are involved in over twice as many crashes as conventional cars.
Another example is the Amazon recruiting software which was developed to eliminate the basis of humans when it comes to employing people. Unfortunately, it failed spectacularly when it was discovered it developed a gender bias. The system was meant to ease the work of evaluating countless resumes. The algorithm had been trained to evaluate candidates on experience. However, since the job positions were mostly in a male-dominated tech field, the machine learning software developed a preference for male candidates. This was due to men having more experience in the field compared to women.
With AI, humanity stands on the brink of greatness. There is so much left to explore, so much left to learn and AI can be the tool to speed up the progress. It can be used to advance humanity at an incredible rate. The only time one will really need to be worried is if they hear about a metal man searching for John Connor. However, AI is still not perfect and is not without its fatal flaws. For it to truly become as powerful as we see in the media, a lot more work is required. As it stands, humans are still vastly superior to it but this does not mean it still can’t be the key we need. Perhaps it won’t be in our life time but who knows what future generations will do with this technology. The future is ready to be explored and wonders to be discovered.