Artificial intelligence is still an uncertain field for the Brazilian government

The concepts about Artificial Intelligence (AI) are from before the existence of the technology that made it possible. Research reveals that studies on the system began during World War II, but it was only in 1943, in an article by Warren McCulloch and Walter Pitts, that neural networks and artificial reasoning structures that imitate the nervous system entered studies.

Faced with the growth of this theme in the world, in Brazil, the discussion reached the Senate, with the analysis of PL 2.338/2023 presented by Senator Rodrigo Pacheco (PSD-MG), who chairs the House, whose proposal is to regulate AI systems in Brazil. The proposed law came from the work of a commission of jurists who analyzed, during the year 2022, proposals related to the subject and legislation on Artificial Intelligence around the world. The commission was chaired by the Minister of the Superior Court of Justice (STJ) Ricardo Villa Bôas Cueva.

In contact with the Em Tempo report, senator Plínio Valério (PSDB) claimed that one of the biggest challenges in establishing a law and regulating this type of system in Brazil is the fact that technology is a field with a high margin for evolution.

“Since it is also a technology in constant evolution, it is difficult to create a framework that will accompany all this development, but it is necessary to establish an environment that obeys Brazilian laws, deals with the rights of the affected people, has transparency and respects the rights authors, for example. It is necessary to have strict regulations on high-risk artificial intelligence systems and on the use of artificial intelligence by public authorities”,

said the senator.

AI entered the agenda in the Senate in 2020, with the creation of a Project to develop a legal framework for the use of Artificial Intelligence in Brazil. Bill 21/20 creates the legal framework for the development and use of AI by public authorities, companies, individuals and various entities, but there is no consensus among specialists.

The author of the project was deputy Eduardo Bismark (PDT-CE), who declared that the objective was to provide the country with legislation that, at the same time, stimulates AI and protects citizens from its misuse. PL had the proposal to create a report on the impact of this technology, with a document prepared by AI agents, with a description of the technology, including risk management and containment measures.

To Em Tempo, the Doctor in Electrical Engineering and professor, associate researcher at the State University of Amazonas (UEA), Ricardo da Silva Barboza, declared that Artificial Intelligence is currently at a stage where the range of possibilities is close to ‘superpowers’. He also warns that technology may now resemble human thinking.

“For many tasks, the result of the AI ​​cannot be differentiated from what a person could provide, in this case the AI ​​passes the Turing test and reaches the human level. Some tasks are impossible for a person, like generating a set of numbers that will win the next contest of the week, and it is impossible for the AI ​​as well. What AI can do is what a person can do, but more accurately or let’s say ‘superpowers’”,

declared the researcher.

Regulation

The possibilities of using artificial intelligence in public authorities are very varied. such as the automation of bureaucratic tasks, storage, research and data analysis. For the vice president of the Education Commission (COED) of the Legislative Assembly of the State of Amazonas (Aleam), state deputy João Luiz (Republicans), this advance can also be a strategy for combating fake news and advancing education.

“We will fight fake news, we will avoid acts of prejudice and incitement to violence in our society. We can further improve our education, an example is our Bill (PL) No. 151/2023, which recognizes the metaverse as a teaching method here in Amazonas. With the new technologies we can advance much more in the areas of health, agriculture, transport, urban mobility and among others, that is, knowledge is power”,

said the deputy.

The metaverse is a virtual ‘world’ that through technologies such as virtual reality, seeks to reproduce augmented reality and the internet. In this collective virtual space there are many possibilities for interaction between its users, such as creating a personal avatar, buying virtual items, playing games and chatting with people.

PL Nº 151/2023, proposed by Deputy João Luiz, aims to establish actions to expand access to technology in the areas of education, training, specialization, inclusion and digital research in the Amazon. The congressman points out that this technology will be greatly exploited in the coming years and that the State must stand out and take the lead with innovative education.

Researcher Dr. Ricardo Barboza states that it would be advisable for the Brazilian academy to be able to serve as a specialist consultant for regulations in the technological area, in order to bring about good results.

“There are certain problems that specialists have been facing for years, and that have already been discussed in congresses and other scientific events. We do not need to repeat mistakes in which there has already been learning and often consensus on the alternatives to be taken”,

said the doctor.

Some of the deputies contacted by Em Tempo said they were not aware of the progress of Bill 2,338/2023 in the Senate, or did not want to talk about it.

PL 2.338/2023

PL 2,338/2023 determines that Artificial Intelligence systems must undergo a preliminary assessment carried out by the suppliers themselves, in order to determine their classification, as being of excessive risk or high risk, based on criteria such as the implementation of the system being in large scale or not; the possibility of causing material or moral damage, irreversible damage or having discriminatory use; the potential negative impact on the exercise of rights and freedoms or the fact that the system affects people from vulnerable groups, such as children, the elderly or people with disabilities, according to information from Agência Senado.

The bill lists systems that will be considered high-end if they are used for activities such as credit scoring; identification of persons; administration of justice; medical diagnoses and procedures; decision-making about access to employment, education, or essential public and private services.

These classifications may be reviewed by competent authorities on a case-by-case basis, to be designated by the Executive Branch.

In the world

Films that portray robots and technologies with an intelligence so advanced that they surpass humans, for many years, as in The Terminator (1984) and Blade Runner (1982). Technological advancement has been talked about for a long time as a ‘futuristic’ possibility, with robots and automatic response systems, but in 2023, it was proven that in reality the future is today

In the world, the field of artificial intelligence development has still been an ‘old west’, according to the report of the MIT Techonology Review, a vehicle that promotes the Massachusetts Institute of Technology, one of the awarded universities in the United States, a reference in ST&I .

In April 2023, European Union parliamentarians requested that world leaders hold a summit to discuss the regulation of AI systems, such as ChastGPT, with the justification that the speed of progress made in the area is being greater than the rulers could foresee. .

In Brazil, AI gained popularity on networks with the discussion of ‘Chat GPT’. The acronym GPT, translating into Portuguese, means Pre-trained Generative Transformer, the artificial intelligence tool has the function of answering questions from users in a free and automated way, based on information and data available on the internet.

The tool, which is still in the testing phase, was launched at the end of 2022 by the OpenAI company, but quickly fell in with the population’s ‘taste’ for performing the most varied functions, from a simple chat to analyzing data and producing texts. The system is easy to access and dynamic, anyone who has access to an internet network can use the tool, which makes the discussion even more complex.

Read more:

Pacheco presents project to regulate artificial intelligence

Artificial intelligence that monitors governments in the Legal Amazon is launched in Manaus