The theme is not new but has never been more present. The concept may not be so simple, and in practice, it is much more complex …
Artificial intelligence can be defined as the ability to give intelligence to machines. Give them the ability to perform activities that were essentially human. To our surprise, these technologies have proven to be more efficient than anyone else. They can already detect breast cancer with 90% accuracy, for example. Autonomous cars, environments that are cleaned up by robots, deliveries with drones and exoskeletons that keep the elderly active are some other examples in which creation beats the creator.
Unfortunately, the machines still cannot think and if they ever reach that level, we will have passed the concept of “Limited Artificial Intelligence”, which is still what we have today. But just like a child, they can learn and every day, they evolve a little, giving enormous facilities to our daily lives. Even what seemed absurdly distant, Artificial Intelligence proves that it may be closer than we thought …
According to Accenture’s annual report (Accenture Technology Vision 2017), seventy-nine per cent of respondents agree that AI will revolutionize how we get information and interact with customers. The truth is that this can already be seen in various companies. The chatbots that begin to replace the attendants are just one example that proves that Artificial Intelligence is gaining maturity. Day by day, AI transforms itself into a new user interface (UI), changing the way we act and how companies act with us.
After a number of advantages, a small problem: making sense of all the information generated by artificial intelligence …. More and more, we can increase the amount of data we have, but we need more effort to interpret and act on them. Data alone doesn’t mean anything and we need to invest efforts to manipulate it. We know that the innovation is necessary, but many companies only remember that innovating can be expensive, and they end up forgetting the investment that this could become …
The application of algorithms to collect data, discover patterns and create accurate automations is called Machine Learning, which is nothing more than a part, nay, techniques of Artificial Intelligence. For example, you’ve probably been scared and wondered how Facebook automatically recognizes the face of your friends when tagging them in some photo … Well, Machine Learning is the answer.
If you ever thought Machine Learning was appalling, then you’ll be even more surprised by Deep Learning, which is nothing more than a deeper machine learning version, based on our own network of neurons. In Deep Learning, huge databases are trained in an unsupervised way. Thus, algorithms are able to identify for themselves images, objects, phrases or words without the need to be pre-programmed by a human.
In fact, both the concept of Deep Learning and Machine Learning are not new. They date back to the 50s or so, which proves the persistence of experts to create a complete Artificial Intelligence, which is not limited and can be able to think, feel and teach. Just like a human being …
