It has been appearing for years in novels and in science fiction movies and series with smart cars, with robots similar to humans, with technologies that do not stop evolving … But artificial intelligence is not only part of cinema and the screen but part of our day a day. It makes any task easier from when the term was coined in the 1950s to today and has many uses and benefits.
In the fifties, neural networks or machines capable of “thinking” by themselves began to be worked on. The term “artificial intelligence” began to be used in 1956 and has been so until today despite the fact that technology and learning have changed a lot. From machine learning to Deep learning: the first uses mathematical algorithms so that machines learn thanks to the data they have entered. As an evolution, in recent decades, Deep Learning has become the most common and also works with algorithms but evolved and improved, it uses neural networks very similar to a real brain.
What is it
Artificial intelligence is a combination of algorithms created in order to act similarly to how the human brain would. A technology capable of simulating that machines can think for themselves. They automate tasks, solve problems, and “think” rationally as a human would. Or in the most similar way possible to this.
According to the European Parliament, artificial intelligence is “the ability of a machine to display the same capabilities as human beings, such as reasoning, learning, creativity and the ability to plan.” According to the Royal Spanish Academy, it is the “Scientific discipline that deals with creating computer programs that carry out operations comparable to those carried out by the human mind, such as learning or logical reasoning.”
Uses and applications
Artificial intelligence is present in our day to day and around us. We use it not only in mobile phones or computers but also in voice assistants that learn our habits and customs to facilitate our day-to-day life or control of home automation, for example. But not only in our home they are useful but also in many other areas such as health, logistics or education.
It is one of the most fashionable uses or the ones that we have talked about the most. Autonomous cars. Autonomous vehicles based on Deep learning that allow you to learn “on the fly” so that the car is intelligent and knows when to accelerate, brake or give way to other vehicles. Although autonomous driving is not yet something we see on the streets, it is ready.
There are several levels : automation such as driving assistance, partial automation, conditional, elevated or full automation. From driving aids (the car does not allow you to change lanes if it detects that there is another car or it brakes automatically if you are going to crash) or high automation that does not require human intervention to move but that allows a person to act in case of that there is a specific failure.
It is one of the most frequent uses that we give to artificial intelligence and is that all mobile phones already have a mode with AI that decides for us . The smartphone detects if you are photographing food, a dog, a person or a landscape. If you are taking a photo during the day, at night, of a monument or a concert.
It will depend on what you are focusing on that the settings and parameters to save the image are one or the other. The AI will decide how it should adjust the brightness, sharpness and all the conditions so that the photo is the best possible. Or even in low-light conditions, some phones take multiple shots when you take one. Before showing it, combine all the ones you have done for a perfect result.
In education, AI has also been used to know what lessons need to be delved further, for example. There are online platforms that use algorithms capable of detecting where each child fails the most and what is doing well. This algorithm adapts the content according to each user and chooses what should appear as an exercise. In addition, they are also working on robots that would help to train in all areas.
Another of the most interesting areas is health. Deep learning algorithms are used in medicine to diagnose diseases, learning from previous cases, from stored data. This is especially useful, for example, when analyzing medical images without the need for a radiologist or to serve as an aid to this person or as a support in developing areas. For example, the iFind project is carried out in several countries (Canada, United Kingdom, Italy…) and its intention is to detect possible birth problems in pregnancies.
There are not only recommendations based on what you do or what you visit, using this data to create profiles about you. Rather, streaming platforms use machine learning to personalize content for each user individually. For example, Netflix does not show its recommendations equally to everyone, but rather is based on our tastes to configure the screen we see when we open the app and to give us one content or another according to this stored data.
- Smart appliances
Vacuum cleaners are the most common case in which we see how AI works. Robot vacuum cleaners learn from our home to be able to move autonomously and vacuum every corner without continuously crashing against walls and doors and without leaving anything dirty. In addition, some more advanced ones are even capable of detecting where there are areas in which you have chosen not to pass and respecting all kinds of delicate objects …
Artificial Intelligence is part of our day to day and it is a mistake to ignore that it is there, so there are already some rules or regulations that speak about these “robots”. Beyond the strictly legal , which we will see in the next few paragraphs, there are many characters and experts who have given their own laws on artificial intelligence, on robotics. From the popular Asimov laws to those given by Satya Nadella, CEO of Microsoft, in a magazine in 2016 and who maintained that it should be designed to help humanity, be transparent, maximize efficiency without distributing dignity, designed to intelligent privacy, with algorithmic responsibility and protected from prejudice in order not to discriminate against people.
The European Union considers that AI has many social benefits in addition to favoring economic growth but considers that its characteristics cannot be ignored taking into account the risks or the safety of citizens. That is why it has a proposal for laws for AI providers or robots or users. They define between several levels according to possible risks and establish different sanctions for infractions or for non-compliance with these rules. Penalties that would go up to thirty million euros or 6% of the total annual turnover.
For this regulation, there is the European Committee for Artificial Intelligence. They explain from the help page of said commission: “The European Committee for Artificial Intelligence will be made up of high-level representatives of the competent national supervisory authorities, the European Data Protection Supervisor and the Commission. Its role will be to facilitate a smooth, efficient and harmonized application of the new AI Regulation.. The Committee will make recommendations and opinions to the Commission on high-risk AI systems and on other relevant aspects related to the effective and uniform application of the new rules. It will also help to generate expertise as a center of competence that national authorities will be able to consult. Finally, it will also support standardization activities in this area. ”
You can stay informed about the AI regulation from the European Union website where you can see the updates of the 2021 plan compared to the 2018 coordinated plan, the changes and differences, the objectives, etc.
Autonomous car legislation
Beyond AI at the European level, there are also controls in autonomous cars and they are adapting. The United Nations Organization signed a regulation in the summer of 2020 that entered into force in January 2021 and whose purpose is to have a series of uniform standards for “Automatic Lane Keeping Systems” or what is the same, level automatic driving. 3 (there are five levels in total) At the moment there are no cars that drive 100% autonomously and are available to users, but this regulation does include those that are partially autonomous.
For example, driving can only be activated where there is no movement of pedestrians and cyclists and where there is a physical barrier to separate one direction of movement from another. There are also rules for manufacturers: all vehicles must have a black box that records all information about driving in automatic mode.