Some time ago, artificial intelligence sounded like something that was only possible in science fiction, but this is no longer the case. Although it's still in its early stages, it's here to stay and revolutionize our lives. More and more industries are incorporating AI into their programs. Some, such as healthcare, finance, security, transportation and education, are employing this type of technology and taking full advantage of its benefits.
For some this may be terrifying, for others it's fascinating, but it's still a reality. That's why we want to help you understand what this means for our future. So you can put any fears about the advance of artificial intelligence behind you and start to enjoy the endless possibilities it has to offer.
What is artificial intelligence?
There's no exact way to define what artificial intelligence is, since the concept of intelligence is quite subjective. However, we could say that AI technology has made it possible for machines to learn from experience, in an almost human-like way. Thanks to this, today we can see how there are systems and machines that adjust to new knowledge and are even capable of performing everyday tasks.
Advances in artificial intelligence have allowed us to see simple things like computers playing chess. Or some more impressive ones, like self-driving cars. This is achieved through deep learning, as this system works with natural language processing.
Despite the great advances seen in recent years, there's still a long way to go with AI technology. It's true that science has managed to develop systems that respond almost intuitively to certain requirements. But we're still a long way from being able to mimic the complexity of how the human mind works.
How does artificial intelligence work?
Artificial intelligence works through a fusion of a huge amount of data with fast, iterative processing that also incorporates intelligent algorithms. This allows the software to auto-learn from the patterns it assimilates.
AI technology is based on a vast field of study, including a large number of theories, methods and technologies. It's therefore important to understand each of them in order to specify how each one works.
Types of Artificial Intelligence
Science is still trying to perfect these technologies to develop more intelligent and intuitive systems. Today we can find four types of artificial intelligence, some of which are already part of our lives and routines. Let's classify and define each of them in order to understand their benefits and their application fields:
The most basic AI systems are of the reactive type. They don't have the ability to remember, so they can't use past experiences as a method of learning. Machines running this type of AI won't have an expanded understanding, as they're designed for something specific and instantaneous.
Remember we previously mentioned the case of computers playing chess? Well, the first computer to achieve this was Deep Blue, created by IBM. It managed to beat the international chess grandmaster, Garry Kasparov. This happened in the late 1990s, and is the perfect example of machines that work with this type of artificial intelligence.
It can assimilate the moves of the match, from identifying the pieces on the board to knowing how to move each one. It can predict the possible moves of the other player and choose a better move to take advantage. But still, it's incapable of having memories.
Deep Blue is only capable of assimilating the rules of chess and focusing on the pieces on the board in real time. It can thus better choose its moves, but it won't acquire any other long-term knowledge.
This model of AI machines, unlike the previous one, can look into the past and assimilate it to make decisions. This type of artificial intelligence is already found in autonomous cars. They also identify and monitor particular objects over time to improve themselves.
This type of fixation is reprogrammed into the memory of these machines so that they can monitor it closely. In the case of autonomous vehicles, traffic light markings, lanes, and other important information such as road curves and obstacles are likely to be included in their system.
As the system assimilates the experiences, it begins to add information such as when to change lanes so as not to collide with another car or hit someone. However, in this type of artificial intelligence, machines can't collect experience over years as a human would. Therefore the information they assimilate is transient. So it won't remain stored in their memory.
Theory of Mind
"Theory of mind" is a term used in psychology to explain the point where not only a representation is created about the world itself, but also about other factors or entities that can affect it. This refers to the fact that people, creatures and even objects, can develop thoughts and emotions that have a determining influence on their own behavior.
This is crucial in our human relationships, as it greatly influences how we relate to others. So, can machines achieve something similar to this? That's precisely what this type of AI is all about. Which comes a little closer to the artificial intelligence we expect to see in the future. Machines with this system are of a more advanced class than previous ones.
These are the machines designed to walk among us, so they are developed to have a better understanding of how we humans think and feel. Therefore, their behavior is adjusted according to what they observe in their environment.
The final step in fully exploiting the potential of AI is to develop systems that can form representations of themselves. This means that developers mustn't only understand consciousness, but also create machines that have it. And of course, this is more complicated.
We humans are self-aware, we know our internal states, and we can predict the feelings of others. The possibility of any technology achieving this feature is quite distant at the moment. But this doesn't mean that studies aren't being carried out.
We may still be far away from self-aware artificial intelligence, but all indications are that we may get there someday. After all, this research is important for understanding human intelligence itself.