Artificial Intelligence, Machine Learning and Deep Learning

By Ronan Byrne

In recent years, much has been said about Artificial Intelligence (AI), Machine Learning (ML) and Deep Learning (DL), being concepts that have become exponentially popular and that, in turn, have generated a lot of confusion among themselves. The purpose of this post is to understand what we mean by each of them and what uses they have in real life.

As a starting point we have to talk about the Artificial Intelligence (AI)understood as the set of technologies that combine algorithms with the objective of creating software that presents the same capabilities as human thought.

Since the concept of artificial intelligence emerged in the 1950s, shortly after the end of the Second World War, it has experienced moments of greatest boom and growth where investments aimed at promoting this technology have been abundant, known as AI summersand other periods in which there has been a reduction in interest and consequently in the investment in funds, known as the winters.

These stoppages were largely due to the fact that the generation of expectations did not become a tangible reality, in many cases due to technical aspects as relevant as the lack of computing necessary for it to be able to develop all the capabilities that were expected. The current reality is that these problems have been solved with new innovations such as the cloud computingwhich give us the adequate storage and computing level to develop its potential, allowing us to overcome these obstacles.

As a result, for a few years now AI has been increasingly present in our day to day in our homesas is the case with Alexa, an Amazon smart speaker that performs tasks that we indicate, such as playing music, turning on the lights or even telling a joke.

The birth of Machine Learning

In the decade of the In the 80s Machine Learning (ML) appearedwhich is a section of Artificial Intelligence, which consists of the the ability of machines to learn by themselves based on data and identification of behaviorswhich allows them to retrain and qualify decisions autonomously, that is, without the need for human supervision.

Trying to simplify its operation a lot, basically a algorithm analyzes a large volume of data, detects behavior patterns and draws conclusions, making the process more efficient automatically.

An example of Machine Learning (ML) in the world of marketing is the personalization of suggestions based on your historyLike Netflix, which recommends series or movies based on your tastes, or Amazon, which positions products on its website or app based on your purchase and browsing history.

Uses of Deep Learning

Finally, we are going to talk about Deep Learning (DL)a subset of Machine Learning (ML), which emerged in the 21st century, and is about algorithms based on artificial neural networks that imitate the functioning of the human brain, being the technique that most closely resembles human learning.

Deep Learning (DL) allows solving complex problems, even if the data set is unstructured and there is no clear relationship between them, the more Deep Learning algorithms are executed and learned, the better the result obtained. This aspect is a differential value with respect to Machine Learning, which requires that the data have been previously structured to offer its maximum performance.

Some examples of practical applications of Deep Learning are machine translations where the algorithms use large volumes of data to train the model allowing the interpretation of languages ​​or, for example, image recognition used in tagging photos on social networks, in fact, spinning with this stitch, One of the great events in the history of Deep Learning was when Andrew Y. NG, in his time at Google, managed to recognize cats among millions of images with a very high level of accuracy.

What is clear is that there is still a long way to go in this area, in fact, the big players technological Market players continue to wage a battle to position themselves as the spearhead of Artificial Intelligence (AI).

Recently, it has been published that Microsoft has incorporated AI into its search engine with the aim of wresting leadership from its main competitor Google. In the words of Microsoft CEO Satya Nadella said that “Artificial intelligence will upend all categories of software, starting with the biggest of all: search.”

The main change they have made in their Bing search engine is the addition of ChatGPT to its search enginean aspect that allows the possibility of asking questions with a high level of complexity and obtaining a specific answer.

This innovation may change the concept of the search engine as we understood it until now when AI solutions are fully integrated, and the most important repercussion, it may shake up the Google advertising model as we understand it until now.