By Luiz Gustavo Pacete
Brazil is among the four countries with the highest willingness to trust AI systems.
The finding is from “Trust in Artificial Intelligence”, a global mapping developed by KPMG consultancy that measures the degree of receptivity of people concerning the technology that has been gaining projection because of OpenAI’s ChatGPT.
A total of 17,193 people in 17 countries were interviewed.
In the general mapping, the study found that 61% of people fear AI systems, while 39% of participants are willing to trust them.
In emerging countries such as Brazil, India, China, and South Africa, trust in technology is higher.
India leads in acceptance, followed by China, South Africa, and Brazil.
The country that comes last in the ranking is Finland.
“The higher trust and acceptance of AI in the Brics group is due to the accelerated adoption of this technology.”
“Trust and acceptance depend on the application: AI use in activities involving healthcare, for example, is the least trusted and accepted,” the study describes.
The study found that most reacted positively when questioning people about the emotions they relate to the topic.
Among them feeling optimistic, enthusiastic, or relaxed about applying the technology.
However, almost half of people also report feeling worried or afraid about AI applications.
“People with positive emotions toward an AI system are likelier to trust AI.”
“Conversely, when people feel negative emotions towards AI, this is associated with lower trust.”
“Further analysis revealed that people commonly experience ambivalent feelings toward AI.”
GREATER ACCEPTANCE AMONG YOUNG PEOPLE
The study identified that younger people, notably Generation Z, and Millennials, trust and accept AI more when compared to older generations.
“These generational effects have occurred in most countries and are particularly pronounced in Australia and the US.”
“In Australia, 25% of older generations trust AI compared to 42% of Generation X and Millennials, and 13% of older generations accept AI compared to 34% of Generation Z and Millennials.”
“In contrast, in South Korea and China, we see a reversal of this pattern, with older generations trusting AI more than younger generations.”
The study showed no significant differences between men, women, and other genders in trust, acceptance, or emotions toward AI.
However, in some countries (the US, Singapore, Israel, and South Korea, in that order), men were more trusting, reporting more positive outcomes.
WHAT IS CHATGPT, AND HOW CAN IT BE USEFUL IN EVERYDAY LIFE
In the last few months, this acronym has appeared in several contents.
And it was not only highlighted when the topic was technology.
It has been mentioned in economics, business, and environment articles.
ChatGPT, just like artificial intelligence, is one of the big topics of 2023.
In this context, it is worth understanding, in practice, what the acronym is all about and, more than this, what impact it may have on business dynamics.
The technology has caught Microsoft’s attention.
Satya Nadella, CEO of the company, has already stated that he wants to ship AI in most of the company’s products.
William Colen, Head of Artificial Intelligence, and Milton Stiilpen, Director of Research & Innovation, both from Take Blip, explain ChatGPT’s usefulness and the reason for its recent resonance.
“GPT, or ‘Pre-Trained Generative Transformer’ technology’, is an Artificial Intelligence (AI) model created from artificial neural networks (machine learning method) and trained with a huge amount of data.”
“Its goal is to understand and produce texts and conversations similar to those done by humans.”
“ChatGPT was an application developed from GPT-3, with specific training and a user-friendly interface, which helps users interact through intelligent and natural dialogues on any subject of human knowledge.”
REPERCUSSION
Stillpen explains:
“The success of this application has suddenly reached the whole world.”
“But it’s the kind of technology that wasn’t built overnight.”
“In 2015, artificial neural networks were already doing something similar, with some early models created through Recurrent Neural Networks (RNN) that simulated the writing of a book author like Shakespeare.”
“The big difference between 2015 and 2023 is that this type of AI model has gained, over the years, more and more capacity to analyze the context of the conversation or the text to be interpreted, summarized, or translated through a mechanism named Attention.”
CHATGPT STRUCTURE
Colen adds:
“The idea behind it is similar to what humans do from childhood.”
“Remember the exercises of filling in the blanks of a sentence given the words around it?”
“Neural network models mimic that mechanism: given the context, what would be the next most likely word? ”
“The recurrent neural networks of 2015 had context capability limited to a few words away.”
“Later created architectures such as Gated Recurrent Units (GRU) and Long Short-Term Memory (LSTM) were able to increase this capability when compared to RNNs.”
“But it was Attention, the mechanism behind Transformers models, that took context capacity, in theory, to infinity if there are enough computational resources.”
“Therefore, models of this kind of architecture can maintain coherence in long sequences as in writing a book.”
“More than that, they can search in their context memory for relations that make possible applications beyond text generation, such as translation, questions and answers, and dialogues.”
With information from Forbes