Why Is Everyone Talking About AI?

Artificial intelligence is a discipline that in recent years has been making its way into the vast world of IT, managing, among other things, to find numerous applications in every aspect. But why is it so important and, above all, why is it spreading so quickly?

To answer these questions, we turn to two further questions: what does artificial intelligence deal with and how does it do it. First of all, let’s try to give a definition, although there is no universally accepted one:

artificial intelligence is a discipline that studies the theoretical foundations, the methodologies and the techniques that make it possible to design systems capable of replicating behaviors that, to a lay observer, would appear to be the exclusive domain of human intelligence.

Thus, if to a newcomer the goal of this discipline may at first glance seem utopian, the way it aims to achieve it in most of its applications seems even more so. In fact, one of the most used methodologies is machine learning, that is, the ability of a machine to learn certain notions autonomously without being explicitly programmed to do so.

Although all this seems rather modern or even futuristic, the idea of having a completely autonomous machine that can fully replicate the behavior of a human being dates back to the mid-20th century. In fact, Alan Turing already proposed then a criterion to determine whether a machine was capable of thinking. That criterion, although it has undergone several modifications over the years, is still used today and is known as the Turing Test.

So, why is artificial intelligence developing only now?

Probably a few years ago the ideal conditions did not exist to carry forward the development of such a complex discipline. Today, thanks to computers with significant computational capacities and to the resulting ability to analyze huge amounts of data in relatively short times, it is possible to envisage a practical outcome for what until a few years ago was only an idea.

Indeed, although the ultimate goal is still far off, there are already many applications that make use of artificial intelligence. Among many, an important role is certainly played by those based on image recognition. Contrary to what one might think, you don’t need to stray from the apps we use daily on our smartphones to find examples. One of them is Facebook, which has been using image recognition for several months now to identify photos in which we appear so as to allow us complete control over content relating to us. In particular, the algorithm is able to recognize, in less than 5 seconds, our face among 800 million photos. The fact that it does so with an accuracy of 98

More generally, image recognition can be used in any field, particularly for detecting certain objects, people and trends in order to keep under control, for example, the goods in a warehouse, a pedestrian crossing or the performance of a listed stock.

That said, to answer the second question, we can certainly think of the rapid development of artificial intelligence as a consequence of the versatility, cross-disciplinary nature and eclecticism that characterize this discipline.

Confirmation of its enormous potential comes from the fact that many IT giants have established their own divisions where they carry forward the development of A.I. One of these is certainly Google, which currently seems to be the company investing the most in this sector, to the point of replacing the name of its old division, Google Research, with Google AI, emphasizing how central artificial intelligence is to its research projects. Not by chance, it is precisely the Mountain View company that created Tensorflow, one of the most used and complete libraries in the field of machine learning.

Consequently, the future for this sector seems bright in every respect: computing power is destined to grow year after year, investments are plentiful and studies arising from research are already bearing fruit.

Having said that, however, we have not considered a fundamental aspect in our whole discussion. Therefore, sliding more toward a philosophical rather than a computer science realm (after all, it is no coincidence that in this discussion the two subjects find a meeting point), I conclude with a question: what if the greatest obstacle on the road to creating a truly conscious machine resides in the human inability to fully understand our own nature? In other words, is man, before the machine, able to conceive what consciousness is?