“These systems do not give correct or incorrect answers”: the challenge of effectively using artificial intelligence

Use the available artificial intelligence tools correctly, and that have taken boom in recent years, first requires understanding how they work to know the way they process information and avoid their misuse, argued the regional director of Microsoft Elevate Skills, Jorge Cella.

“There is a huge challenge in all this and that these tools, as well as they are powerful and simple to use, require training, require that I learn to use it. What in principle does not seem because it is a chat: I ask questions, give me answers. But they have to teach me to ask questions, and the reason for this has to do with the skills that are needed to get all the value back“, held.

In the last five years, the availability of generative artificial intelligence tools has increased, above all, in “chatbot” programming that simplify their use. The Technological Executive explained that the platforms on which users write a question and receive answers- as chatgpt- have massified the use of artificial intelligence, because it is no longer necessary to know technical language to make the tool work and there are multiple free options.

The increase in the use of “chatbots” has proven to have its risks. The death of suicide of a 29 -year -old American, earlier this year, led his mother to discover that he had been consulting a chatbot instead of a mental health professional; The note he left when he died was written by technology, he reported The New York Times. A Florida mother sued a technology company last year because she considered that a chatbot was responsible for the death of her 14 -year -old son, CNN published. The Wall Street Journal He reported, in August, that a man killed his mother in Connecticut and then deprived himself of life. Subsequently, evidence was found that he felt persecuted and frequently used an artificial intelligence platform that validated and encouraged his paranoia.

Cella said it is not the first time that people replace communication or human contact after the arrival of new technologies. With the emergence of search engines, such as Google, he recalled that there were cases of individuals who used the platform as confessionals and took “the first response as a reality.”

“We are very concerned about that, and that is why the learning of use is so imperative. Once they learn and understand how technology works according to their own thinking to take the answer. And you can ask personal questions, critically, and so you understand where the answer comes from,” he said.

“The way to protect and protect others is to explain how it works. The demystification of technology is so important because, if not, we will have consequences for their misuse, and those cases make me make wrong decisions, especially when rest in which the decision is taking a system that I have no idea what the data is or what is the process that follows that I build an answer,” he added.

The Microsoft executive stressed that there are alliances to support educators to train in the management of these platforms. The company has agreements, in the United States, with the American Federation of Teachers (AFT) and the National Education Association (NEA). With the latter, Microsoft will provide training and certifications to more than 10,000 American educators and 500 leaders of the educational industry this schoolthe company reported in written statements.

So far this four -year At the legislative assembly, at least 17 measures have been presented Related to artificial intelligence, focused on issues such as education, security and politics, among others.

Cella stressed that it is necessary that the users of these tools understand that they are based on collecting information, according to their programming.

“I always like to explain that These systems do not give correct or incorrect answers, give the most likely answer according to the data they have. So that system, I don’t know what data, I don’t know how you found that probable answer, so you have to be careful and understand what is the most important ability to use these tools, which is critical thinking. It is not technological, it is to question the answers, to be able to put my own thought above any type of response that gives me, look for other sources of information, etc., ”he said.

Given this, Cella stressed that Generative artificial intelligence should be seen as a productivity tool. He argued that artificial intelligence alone will not eliminate jobs, but will create new sources of work and those who are in the labor market must be trained for them.

He stressed that, through its ELET SKILLS program, Microsoft offers training, free and available in Spanish, on generative artificial intelligence and how to use various tools correctly. In Latin America, more than 16 million people have begun to take courses and about 5 million have completed them. In Puerto Rico, about 122,000 people have begun to take courses since July 2024, and around 62,000 have obtained a certification when completing them.