The Internet of Things (IoT) is part of the so-called Web 3.0. IoT is the ability to make a physical object connected to the Internet able to interact with people and be connected to other physical objects to perform tasks. The internet of things is based on machine learning and artificial intelligence. The term first came to public attention in 1999 in a lecture by British technology expert Kevin Ashton. The initial idea was related to the aim of helping computers to know things through data collected without human intervention. Technology has evolved and the Internet of Things means more than just knowing things: machines are already able to perform actions. One of the main components of the Internet of Things is sensors. They capture external information and send it to the system that is connected to the internet so that a particular object performs an action. For example: an audio sensor is able to hear the user’s order to turn on the light. The sensor captures the information and passes it to the house lighting system, both of which are connected to the internet. Soon the lights are turned on. Smart homes, smart cars, and smart cities are terms used to designate the consequences of the application of IoT in everyday life. With the implementation of 5G internet, connections will become faster, enabling more linking of objects to the internet, expanding the IoT.

ery simply, the term journalism of things (JoT) joins the concepts of journalism and the internet of things, as formulated by Brazilian researcher Marcelo Barcelos in his PhD thesis, and also by Jakob Vicari in his book Journalismus der dinge, both from 2019. With the development of technologies based on artificial intelligence, automated systems are already capable of producing journalistic content. News agencies and media outlets already use AI and algorithm-based tools to produce content. These tools search the internet for data, cross-reference the information, and produce articles. For trivial hard news coverage like road accidents, traffic information, and weather events, the use of these AI-based tools for news production can be effective in answering the five W’s (who, what, where, when, and why), which are the basic elements for writing a journalistic text. Technology can become an ally of the newsrooms since it frees the journalist to do jobs that require more time and dedication to the investigation of more complex facts and that require a more critical perspective. But it also raises new debates, given the impact of this technology on the profession and users.

There interesting projects going on at WDR; in an interview to AXSemantics, Vicari speaks about a project called Super Cows, where three dairy cows “reported from their lives”. Sensor data gave the animals a voice. And with #bienenlive, the WDR spectators were taken live into the life of bees. Explained in a very simple way, in this project a simple text robot translated the data into a diary of the queen bee. This new world and its challenges bring to the front the discussion of the ethics surrounding the internet of things, the hybridization/robotization of human affectivity, the isolation and harassment of intelligent machines, amongst other things.
 

Natural Language Understanding is a subtopic of Natural Language Processing in artificial intelligence that deals with machine reading comprehension. It involves modelling human reading comprehension, or, in other words, analysing and translating the input according to natural language principles. Smart speakers, for example, use Natural Language Understanding to be able to respond to the user’s request or command via voice. Systems that use writing, such as chatbots, can also use an NLU-based system. This tool can understand the user’s intention and not just decode words. For example: when a person is using an online restaurant service and asks what flavours of a certain dish are available, the machine is able to interpret that the question has to do with the menu and deliver that option to the customer. But it can only reply with scripted answers. In a simpler system, the customer would have to ask for a menu in order to have his or her request fulfilled. NLU is therefore a tool that approaches human reasoning by reading the intent behind a sentence. Natural Language Understanding works in conjunction with Natural Language Generation, which is the ability of a machine to generate natural language, spoken or written, in response to a user. Together they form a Natural Language Processing system.

Artificial intelligence is defined as the ability of a digital computing device, or a robot controlled by a computer, to perform tasks commonly associated with intelligent beings.  Since the development of the digital computer in the 1940’s, it has been demonstrated that computers can be programmed to carry out very complex tasks – such as discovering proofs for mathematical theorems or playing chess – with great proficiency.

One of the key moments for AI was the publication of Alan Turing’s Computing Machinery and Intelligence in 1950. In this paper, Turing explored the idea of how to determine whether machines can think. Although the development of AI systems has progressed, and despite the continuing advances in the field, none of the projects developed to date can yet match the flexibility of humans in a wide range of domains. Nonetheless, some computers/programs have attained the performance levels of human experts and professionals when it comes to doing specific tasks, and the complexity of the tasks that these programs can manage is growing. These applications range from medical diagnosis (under very specific conditions), computer search engines, and voice or handwriting recognition.

AI is not just something we see in movies, it is already part of our daily practices, for instance every time we use Siri or Alexa, or when we use a search engine such as Google, or social media platforms such as Facebook, or even online banking. There are also less recognizable uses for AI, like behavioural algorithms, suggestive searches, autonomously powered self-driving vehicles, and automated news production in some newsrooms.

Robot journalism, also known as algorithmic journalism, automated journalism, or automated reporting, consists in the production of news stories/articles by computer programs. In essence, it is the use of Artificial Intelligence (AI) applied to journalism practices. Through AI software, stories are produced automatically by computers rather than by human reporters.

These programs interpret, organize, and present data in human-readable ways. The process involves algorithms that scan large amounts of data, select from an assortment of pre-programmed article structures, order key points, and insert details such as names, places, amounts, rankings, statistics, and other figures. The output can also be customized to fit a certain voice, tone, or style.

Automated reporting is possible for events that happen regularly and the results of which can be quantified. “Quantified”, meaning that there is structured data available about the outcome of the event. A few examples are sports matches, stock prices, quarterly company reports, frequently published socioeconomic statistics, and election results. Investigative reporters can use AI to make sense of huge document leaks, geographical data, or satellite data. It can also be used by video editors to help organize hours of video footage. The use of AI in journalism is already a reality and some newsrooms are using it for some of the mentioned examples, although in some cases articles or other journalistic products created by AI benefit from human contextualization by referring, for example, to the causes of the presented facts.

According to Techopedia, “in its purest sense, an algorithm is a mathematical process to solve a problem using a finite number of steps. In the world of computers, an algorithm is the set of instructions that defines not just what needs to be done but how to do it”. So, in brief, an algorithm is a set of instructions designed to perform a specific task. 

Algorithms are used in very diverse ways. Their development is crucial in computer science. They are a key part of artificial intelligence, databases, graphics, networking, operating systems, and security, just to mention some.

Sometimes, when we talk about algorithms, we tend to think of them as programming tasks, but algorithm development is more than that. To develop an algorithm, it’s necessary to understand all the alternatives available, ranging from hardware, networking and programming language to performance constraints. It also requires understanding what it means for an algorithm to be “correct” in the sense that it fully and efficiently solves the problem.

To understand what an algorithm is and how it works, in a great deal of literature we find a very simple example: a cooking recipe, as we set and follow several rules that have to be followed to achieve the desired outcome.

There are several kinds of algorithms, and in a very simple way we can divide them into four types: heuristic, metaheuristic, machine learning and statistical.

Regarding the media, we are especially interested in machine learning algorithms, which can, for example, liberate journalists from dull and more repetitive tasks, such as analyzing big amounts of data. Some of their uses for image recognition include identifying an object as a digital image, based on the intensity of pixels in black and white images or colour images.

Another use is the translation of speech into text. Certain software applications can convert live voice and recorded speech into a text file, but also can be used in voice activated assistants. Machine learning can also be used to extract structured information from unstructured data. A machine learning algorithm automates the process of annotating datasets for predictive analytics tools, which can be used by data journalists.

Artificial Intelligence (AI) is embedded in everyday life, business, government, medicine and more. In order to trust these systems, it is of great importance that they respect ethical principles. As stated by IBM (one of the oldest IT companies, and one that has a vast history in working with AI), only by embedding ethical principles into AI applications and processes can we build trustworthy systems. The importance of ethics in the Artificial Intelligence field is widely recognized, and on 24 November 2021 UNESCO’s General Conference, during its 41st session, adopted the Recommendation on the Ethics of Artificial Intelligence. The main areas that are debated as regards AI ethics are: privacy and surveillance; manipulation of behaviour; opacity of AI systems; bias in decision systems; human-robot interaction; automation and employment; autonomous systems; machine ethics; artificial moral agents; and singularity. For example, IBM grounds its approach to AI ethics on three basic principles: (1) the purpose of AI is to augment human intelligence; (2) data and insights belong to their creators; (3) new technology, including AI systems, must be transparent and explainable.

Looking at the use of AI in news organizations, we see that transparency is a core issue when it comes to AI ethics. Algorithm transparency and accountability is an important discussion point. It is not easy to discuss AI boundaries, even though this is not a new subject, as they already were a topic of debate in the last century. While at that time AI applications were more limited to academic and technological environments, today AI is embedded in our daily lives, making the subject more important each day.

Machine learning is a branch of artificial intelligence and computer science which focuses on the use of data and algorithms to imitate the way that humans learn. Nowadays machine learning is probably used by the average citizen several times a day, as it can be found applied in web searching, practical speech recognition, weather prediction, self-driving cars, and so on.

In 1950, Alan Turing created the Turing Test, theobjective of which is to determine if a computer has real intelligence. The basic premise is that to pass this test, a computer must be able to fool a human into believing it is also human. In the history of AI, machine learning has evolved from that time to today.

Machine learning falls into three primary categories: supervised machine learning, unsupervised machine learning, and semi-supervised learning. Although machine learning represents a growing and important field with many gains in several areas, it also raises ethical concerns, such as its impact on job creation/loss, privacy, bias and discrimination, and accountability. In journalism, some of the areas of implementation of machine learning are: information search, content generation, and recognition systems.

Very simply, we can say that robotics is an interdisciplinary field of science and engineering dedicated to the design, construction and use of mechanical robots. As defined by Encyclopaedia Britannica, robots are widely used in such industries as automobile manufacture to perform simple repetitive tasks, and in industries where work must be performed in environments hazardous to humans. Besides the technical disciplines of robotics, ethics is important, as some of the work in the field raises ethical questions, as we see, for example, stated in Asimov’s three laws of robotics. Meanwhile, there continues to be some confusion about the terms artificial intelligence (AI) and robotics, as they are two separate fields of technology and engineering. Nonetheless, they are often combined, such that artificial intelligence and machine learning are also involved in robotics. In recent years there has been overlap between these fields, causing some confusion as to what can be described as a physical robot and a software robot. To prevent confusion, the word “bot” is used to describe software robots that have no mechanical body. In journalism, bots are used most often, and since they don’t have a physical body, they are software robots, backed with Artificial Intelligence and machine learning.

Big data is a term that refers to data that is so large, fast, or complex that it’s very difficult to process and manage using traditional methods, and also challenging storage capacity. Accessing and storing large amounts of information has been possible for a long time, but the concept of big data gained in importance in the early 2000s when analyst Doug Laney articulated the now-mainstream definition of big data as the three V’s: volume, velocity and variety. Volume represents the big amount of data that is being collected and stored by several organizations, from a great variety of sources, like smart (IoT) devices, industrial equipment, videos, images, audio, social media and many more. Velocity refers to the fact that data streams into the organizations at an unprecedented speed, coming from sensors, RFID tags or even social media, bringing with it the challenge of managing flows in a timely manner. Variety means that data comes in different formats – for example it can be structured or unstructured, it can be numeric data in a database, and it can be text documents and videos, and this also brings challenges to the use of this data. There is an enormous and unprecedented growth in the collection and analysis of digital trace information regarding human and natural activity, such as localization information from mobile phones or pollution sensors. If we think about networked communication, smart cities, and the Internet of Things we see how big of an impact this can have in our world. In addition to their impact on technology and society, these developments have particular significance for the media industry and for journalism as a practice and a profession. Although there is still much to learn about these impacts, we can surely say the use of algorithms in the newsroom, or the emergence of automated journalism, are some of the uses of big data in media organizations that will have and are already having impacts in this field.