It all started back in World War 2 with Alan Turing breaking the Enigma code, which was a machine the German Nazi’s used for all their communication’s.Alan Turing is believed to be the father of Artificial Intelligence. You can learn about who Turing was and all his life’s work in an article I wrote about him Here.
People are stating that American computer and cognitive scientist John McCarthy was the real father of AI, since he was the one that invented the term itself, but if u read the 1950’s paper of Computing Machinery and Intelligence you can clearly see that Alan Turing was the founding father with the creation of the Turing Test.
Turing in 1950 created the framework of the Turing Machine in a form of a test, which is a test that determines the ability of the machine to exhibit intelligent behavior equivalent or indistinguishable to that of a human.
In 1956 the Dartmouth Summer Research Project on Artificial Intelligence was introduced, thus creating the first AI research workshop. With researchers from MIT, CMU and IBM meeting together to create this AI research, everyone thought that in the very near future, machines will be capable to do any work that a human was able to, thus replacing us.
Even though they made huge progress for couple of years, after the 1960’s everything slowed down, because no one had foreseen how difficult it would be to actually create a true AI.
In sort term a true AI system will be able to think like a human can. Now if you are not in a field like Computer Science or equivalent you can’t grasp of how difficult of a task this truly is.
From the birth of AI, the whole idea was that you take a computer software and slowly you can teach it to do something, you can think of it like a small baby learning to walk. The system is starting from a point of not knowing anything and slowly with the mistakes it makes, it can learn and adapt until the outcome is achieved.
A small example it would be the texting pattern in your smartphone. When you first buy it, the prediction on the text you are typing is basic to say the list, but as time goes by it learns and adapts to your texting pattern, thus giving you predictions of next words or even next sentences, you are more likely to use.
Another example of AI is speech and image recognition. In the last couple of years most users are getting more familiar with talking to their computer or smartphones rather than using the keyboard. The big giants like Google, Facebook, Microsoft are constantly making small improvements to all these technics every single day.
Now all of these advancements became possible by implementing the use of Neural Networks with an AI System. In essence true AI isn’t possible without the use of a neural network, they are kind of the brain of an AI system.
Neural Networks aren’t something new, they were also created back in 1950’s. In its simplest form it’s an algorithm that is able to use huge amounts of information to make the system learn and evolve. This whole process is referred to as deep learning and it’s the foundation of mostly anything that is being created today. Even system’s that were created back in the day like Watson, are now augmented with deep learning.
What most people don’t realize is that all of this kind of breakthroughs are not possible by a human being. No programmer can create a system like that, in reality programmers give the system a learning algorithm, then open it to insane amounts of data, thus making the system able to learn and figure out for itself what is what.
From the beginning of human life, all we are trying to accomplish is to improve and become better. Even though we have achieved so many things there are still huge problems that are facing the human race.
To be able to understand why a computer will be better to figure out the solution to problems like this, I have to explain to you the basics of how a computer thinks or works.
If you take out some basic factors in a system, a computer can calculate something based on the speed the CPU clock runs at. Most CPU this days are in the range of 1,5 to 5 GHZ. The measure GHZ is referring to the frequency, and frequency is the reciprocal of time. All this means, is that the number gives the amount of work being done per unit time. Let’s say a CPU runs at 3.5GHZ, this means that it can produce 3,5 billion cycles per second. So, in a sense it can make a prediction based on 3,5 billion thoughts per second.
So, a machine can think a lot faster, with no limitation of how many parameters you through at it.
But, like any technology ever created there are always two sides.
Human’s don’t like, nor are good at repetitive tasks. A machine can perform a single task hundreds of millions of times without ever stopping.
Other than the simple everyday tasks that can improve our life like Traffic Controllers, Language Translation, intention recognition, chatbots and thousand other cases, the biggest field an AI can be used is diseases and also answering some of the biggest problems we are facing as a human race, one of them being over population.
Even if a technology is created for something specific that can save 1000 life’s, there will always be someone that can use the same type of technology to destroy life’s.
That can probably be said for mostly anything human race has created, but in the field of AI it’s not that simple, a pure AI can truly take over the world.
Most people think that movies like the terminator with Skynet and Matrix are just fiction. If we are not very careful about the creation of an AI system, those scenarios are more than possible.
Elon Musk which is a world known engineer and investor, in the last couple of years has bought shares in some of the known companies that focus on advancing AI, just to be able to keep an eye on what its been done, so he can prevent the worst-case scenario. He has also in numerous occasions stated that we should be very careful about AI.
In my opinion, Artificial Intelligence can truly change the way we live, and I am more than certain it will achieve it in the future.
Up until now no machine has past the Turing Test, so we still have a lot of improvement and research ahead of us until a true AI gets build.
But when it comes to the biggest downside of a true AI system, what I think is simple.
A smarter person than you, is able to manipulate you in some way or another. A sentient machine within minutes of internet access will be smarter than any human that ever lived combined. It will have access to all our history, all that we have done and especially everything that we have done to each other, and the planet.
This is kind of the scary aspect, but if it where up to you given all our history, wars, environmental pollution etc, what would you do, destroy the humans race so the planet can be saved or not?
“I visualize a time when we will be to robots what dogs are to humans, and I’m rooting for the machines.”
- Claude Shannon