ChatGPT is the new trend of the technological world and it adds millions of lovers and some detractors. Basically it’s a program artificial intelligence (better known by the initials AI of artificial intelligence), that is, the ability of a machine to reproduce the skills of human intelligence, namely its capacity for reasoning, learning and creativity.
This means that an electronic device equipped with AI systems (software and algorithms) will be able to function in a human reasoning logic, being able to analyze, learn and decide on different situations and/or problems.
A device with AI receives information (eg data, images, sounds), processes it and responds according to what is asked of it, and can even adapt its behavior according to the effects caused by its previous responses.
The device itself can be constructed in such a way as to have its own sensors (eg camera) to record new data according to the programming defined to do so.
Digital AI technologies are not new to this century. They have been working on for more than 50 years, but the evolution of recent years has allowed for enormous progress, namely in the processing capacity of these systems for ever-increasing amounts of data.
A famous witness to this evolution is ChatGPT, a chatbot made available to the public in November 2022 by OpenAI, an American non-profit company that is part of a laboratory specializing in research in the field of artificial intelligence.
Artificial intelligence: what is ChatGPT?
A chatbot is a program that seeks to replicate the responses of a human being during a conversation between people.
You have certainly come across some of these programs that have quickly become famous in recent years as virtual customer support assistants, via online or telephone conversation.
ChatGPT is why a tool that processes human natural language and which was programmed to answer questions on a wide range of topics.
This chatbot is a kind of language machine that uses statistics from reinforcement and supervised learning to display words, phrases and comments in your answers.
He doesn’t exactly know what the words mean, but he knows how they are used and uses that knowledge to organize and reproduce them in his answers.
The novelty of ChatGPT, compared to other AI chatbots, is that it responds more efficiently and is able to summarize large amounts of information which allows it to be used to program code, answer exam questions, write articles, create automatic questionnaires about specific courses or create news.
You can even imitate writing styles and can improve your performance as you receive more questions. This effectiveness results from the fact that this system is trained not only in accessing the Internet, but also in accessing more complete and better validated databases.
ChatGPT is already at version GPT-3.5 and its performance is proving to be so impressive that it has already been dubbed a true “information war machine”.
Artificial intelligence: the downside
If, on the one hand, engineers and businessmen see the use of ChatGPT as a precursor to a frontier that must be crossed in favor of increasing productivity and creativity that will allow the development of new horizons and solutions for current problems, on the other hand, scientists social media and journalists who raise legitimate and growing concerns.
What can crash in ChatGPT?
The weakest point of these artificial intelligence systems is the incorrect answers they can provide to their users.
A chatbot that uses the Internet has access to a huge base of information, but we all know, from our own research experience, how much of that information can be incorrect and incomplete. Even for a system that learns through use, this risk will always be real.
Another important aspect is the fact that, even though ChatGPT does not produce incorrect answers, it can replicate in these responses the potential bias of its algorithms on economic or social issues (race, gender, people and social classes or countries disadvantaged by digital inequality, etc. ).
Since the Internet is a broth of content from marketing, self-promotion, news and opinion, it is necessary that systems of this type of artificial intelligence depend not only on the quality of the data and information they use to validate their answers, but also the algorithms they use to constantly check the reliability of the information they access so as not to become biased or agents of misinformation.
Impact on the future of work: the tipping point
This type of chatbot makes it possible to automate boring and time-consuming tasks (researching data, writing reports or meeting minutes, etc.), helping to improve the productivity of more qualified workers by freeing them up for more creative tasks.
Its maximum performance and effectiveness is expected to happen in a closed circuit, in the face of a specific theme/technique.
But on this evolutionary path, there are many professions and jobs that are going to change or disappear, especially among the less qualified, without there being time to find a replacement in the world of work.
With regard to the ChatGPT example alone, analysts predict that among editors, journalists, analysts, customer service agents, control and quality engineers, about 8% will be directly affected.
These are professions that, even if they do not disappear immediately or definitively, will require a great deal of adaptation on the part of these professionals. The requirements will necessarily involve progressively abandoning the more automated tasks to dedicate themselves to more complete tasks or those committed to innovation, but these adjustments normally take longer to materialize than the adoption of new technologies.
In this sense, if we want to take advantage of the economic and evolutionary gains offered by artificial intelligence to humanity, it will be necessary to reduce its negative impact on workers and existing social structures.
Impact of Artificial Intelligence on the job market