This stuff is AI

The tremendous success of Deep Learning has created such a hype (and wave) in science and technology that has resulted in the "this/everything has/is AI" fallacy.?Something that is terribly wrong, misleading, and creating loads of confusion among professionals and, more worryingly, the general public. As a result, we get the counterargument "that/nothing isn't/has AI".?I believe this debate is pointless, as rooted in fallacies and a fundamental misconception of what the term AI really means. Let me explain myself in the following paragraphs.?

Artificial Intelligence or AI is an academic discipline, part of Computer Science, and officially founded with that very same name in 1956. Actually, one decade earlier than Software Engineering (1968-9). AI roots date back (at least) to 1943 (with McCulloch and Pitts' first artificial neuron model) and 1950 (with Turing's computing machinery and intelligence, and so the famous Turing Test). Nevertheless, many other AWESOME scientists contributed to setting strong philosophical and scientific foundations to such a discipline, from the 1940s (remember Norbert Weiner’s Cybernetics and Claude Shannon’s Information Theory) all the way until today. It is a story of a scientific endeavor that has been as magnificent as difficult, aiming to discover how human intelligence could possibly happen. It reminds me of other awesome scientific endeavors: to harness the power of the atom, or that of the genes as the building blocks of life. No more no less.?

As in any human endeavor of such scale, many, many theories have emerged, within many different paradigms and approaches, to accomplish the mission. One of them is Connectionism (or Artificial Neural Networks), rooted back to McCulloch and Pitts' artificial neuron model. This is currently well known as Deep Learning.?So, making my numbers, for almost 80 years AI has been a prolific scientific discipline where Connectionism has given rise to modern Deep Learning (DL). Quite a long journey, isn't it??

It is interesting to remember that AI was so despised by the rest of the Computer Science (not to mention the rest of the scientific community as a whole) because of its fantastic goal, that other important disciplines, highly related, were founded independently to avoid the stigma. This is the case of Machine Learning - in the intersection of Maths and Computer Science; also of Computer Vision, a very popular and dignified discipline to work in since the 1980s. But listen: Connectionism was not well accepted within their walled gardens! Those crazy scientists working from postulates and approaches rooted in that crazy idea of AI!!?

The story reminds me of the "Only one small village of indomitable Gauls still holds out against the invaders." The so-called Godfathers of Deep Learning got tired at some point and the rest is history: they fought back and took Machine Learning, Computer Vision, and the whole Computer Science by storm (an oversimplification of a fascinating story, by the way).

Paradoxes of life, now, everything that happens (as research and innovation) in all those disciplines, and beyond, is not only Deep Learning but AI!

In other words, the success of Deep Learning has created hype resulting in the "this/everything has/is AI" idea; when people are just using some sort of Deep Learning algorithm - or not. And this is fought back by people saying "that/nothing isn't/has AI"; claiming it is not "really intelligent", or "human intelligence", or "real intelligence" or "not as intelligent as a human being", or something of the sort, whatever it means.?

So, I strongly suggest to both sides to stop using the term AI unless they are talking about the scientific discipline. Actually, stop using the term AI altogether when not in a scientific context. Use the term Deep Learning instead; it is fancy enough, don't you think so? As a direct benefit, we would stop wasting our mental energy thinking about whether that algorithm is AI or not. That is an academic debate, no need to move it into the general public.

IMHO, the general public must be aware and worried about other problems instead: algorithmic bias and injustice, lack of privacy, personalized control and surveillance, autonomous weapons and policing, etc. those Deep Learning algorithms shall bring about.?